Tamarack Defence Systems
Every technology as a list
Tamarack Defence Systems is a fictional composite, a Canadian defence supplier invented for this demonstration. The radar reads how ready the firm is for AI across eight parts of the organisation, from the boardroom to the shop floor. Every dot is an AI practice: most are already proven in allied defence, and the few near the centre are what the allies are still working out. The gap view, under Readiness, lights the practices the firm is not yet ready for. What the radar shows is where Tamarack has decided, where it has not, and which decisions would move it furthest.
About this assessment Assessed by Trailblazed Ventures, with the firm's executive team on October 21, 2026.
How ready the organisation is
What is scored. Each family is scored on three dimensions, each of which answers one question the survey asks:
- Regulatory: do the rules permit these technologies here, and how much approval do they take?
- Acceptance: will staff, the public and oversight bodies accept these technologies?
- Capability: can the organisation’s skills, processes and systems absorb these technologies?
How it is scored. Each dimension is scored at one of five levels, from Beginning (1 of 5) to Mastering (5 of 5). Starting a pilot of a technology needs every dimension at Learning (3 of 5) or above; adopting it across the organisation needs every dimension at Developing (4 of 5) or above, so that it runs on a set process rather than on specialists.
What the five levels mean
- Beginning (1 of 5): nothing is in place yet, and nobody owns it.
- Emerging (2 of 5): the question has been looked at, but there is no set route, plan or budget.
- Learning (3 of 5): it is done in some areas, led by specialists.
- Developing (4 of 5): a set process runs everywhere, though results still vary.
- Mastering (5 of 5): it is routine everywhere, and it is reviewed and improved on a schedule.
For Leadership & Strategy, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5).
Your organisation is not yet ready for any of this family's 7 technologies, all of which are in the Pilot or Adopt rings.
AI is in the annual plan as a paragraph. The board has discussed it twice and decided nothing; four vice-presidents each own a piece. The rules would allow far more than the firm has chosen to do.
- Its weakest dimension is Capability, scored at Beginning (1 of 5).
- Raising Acceptance and Capability to Learning (3 of 5) would make your organisation ready to pilot these technologies.
- Raising all three dimensions to Developing (4 of 5) would make your organisation ready to adopt them.
- At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget.
- Regulatory is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Regulatory is at Learning (3 of 5), which means that a route to approval has been used in some areas, led by specialists, and some coming rule changes are known.
- Acceptance is scored at Emerging (2 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Acceptance is at Emerging (2 of 5), which means that views are known for a few high-profile uses, but there is no plan to answer concerns.
- Capability is scored at Beginning (1 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Capability is at Beginning (1 of 5), which means that no one owns these technologies, and the skills, processes and systems they need are not in place.
This family's technologies, ring by ring
- Adopt ring, 5 technologies. These are proven in operation, but your organisation is not yet ready to adopt them.
- Pilot ring, 2 technologies. These are real enough to test, but your organisation is not yet ready to pilot them.
For Governance, Ethics & Assurance, overall readiness is 3.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Learning (3 of 5).
Your organisation is not yet ready for any of this family's 6 technologies in the Pilot or Adopt rings.
The compliance culture is real: principles drafted, a working group meeting, impact assessment understood. What is missing is authority, so the working group advises and nothing gets approved.
- Its weakest dimension is Capability, scored at Learning (3 of 5).
- Raising Capability one level, to Developing (4 of 5), would make your organisation ready to adopt these technologies.
- At Developing, Capability looks like this: senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary.
- Regulatory is scored at Developing (4 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Regulatory is at Developing (4 of 5), which means that approval follows a set process everywhere, and coming rule changes are planned for, though results still vary.
- Acceptance is scored at Developing (4 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Acceptance is at Developing (4 of 5), which means that those affected are engaged everywhere by a set process before each use, though support still varies.
- Capability is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Capability is at Learning (3 of 5), which means that leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help.
This family's technologies, ring by ring
- Adopt ring, 6 technologies. These are proven in operation, but your organisation is not yet ready to adopt them.
- Scan ring, 1 technology. It is at an early stage, so there is nothing to be ready for yet.
For Security, Compliance & Export Control, overall readiness is 4.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Developing (4 of 5).
Your organisation is ready for all 6 of this family's technologies in the Pilot or Adopt rings.
The firm's strength. Export control, cyber and classification discipline are mature and trusted by the customer. Security has already blocked what it should and is ready to enable what it could.
- At Mastering, Regulatory looks like this: approval is routine and the same everywhere, and each rule change in view has a plan.
- Regulatory is scored at Developing (4 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Regulatory is at Developing (4 of 5), which means that approval follows a set process everywhere, and coming rule changes are planned for, though results still vary.
- Acceptance is scored at Mastering (5 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Acceptance is at Mastering (5 of 5), which means that no significant objection is outstanding, and those affected can see and question how these technologies are run.
- Capability is scored at Developing (4 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Capability is at Developing (4 of 5), which means that senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary.
This family's technologies, ring by ring
- Adopt ring, 5 technologies. These are proven in operation, and your organisation is ready to adopt them.
- Pilot ring, 1 technology. It is real enough to test, and your organisation is ready to pilot it.
- Scan ring, 1 technology. It is at an early stage, so there is nothing to be ready for yet.
For Data & Digital Foundations, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5).
Your organisation is not yet ready for any of this family's 7 technologies, all of which are in the Pilot or Adopt rings.
Three data estates and owners only for engineering data. Three pilots stalled on data they could not get. The skills to build a platform are thin; the mandate to build one does not exist.
- Its weakest dimension is Capability, scored at Beginning (1 of 5).
- Raising Regulatory and Capability to Learning (3 of 5) would make your organisation ready to pilot these technologies.
- Raising all three dimensions to Developing (4 of 5) would make your organisation ready to adopt them.
- At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget.
- Regulatory is scored at Emerging (2 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Regulatory is at Emerging (2 of 5), which means that the rules have been checked, and approval is possible case by case, with no set route.
- Acceptance is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Acceptance is at Learning (3 of 5), which means that in some parts of the organisation, those affected are consulted and their concerns acted on, led by specialists.
- Capability is scored at Beginning (1 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Capability is at Beginning (1 of 5), which means that no one owns these technologies, and the skills, processes and systems they need are not in place.
This family's technologies, ring by ring
- Adopt ring, 4 technologies. These are proven in operation, but your organisation is not yet ready to adopt them.
- Pilot ring, 3 technologies. These are real enough to test, but your organisation is not yet ready to pilot them.
For Engineering & Product, overall readiness is 3.0 of 5, with all three dimensions scored at Learning (3 of 5).
Your organisation is ready for 2 of this family's 6 technologies in the Pilot or Adopt rings.
Strong where the products are strongest: the naval sensor team tests models well and is ahead of the firm's governance. Elsewhere, engineers want tools the firm cannot yet permit.
- All three dimensions are scored at Learning (3 of 5).
- Your organisation is already ready to pilot these technologies.
- Raising them one level, to Developing (4 of 5), would make your organisation ready to adopt them.
- At Developing, Regulatory looks like this: approval follows a set process everywhere, and coming rule changes are planned for, though results still vary.
- Regulatory is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Regulatory is at Learning (3 of 5), which means that a route to approval has been used in some areas, led by specialists, and some coming rule changes are known.
- Acceptance is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Acceptance is at Learning (3 of 5), which means that in some parts of the organisation, those affected are consulted and their concerns acted on, led by specialists.
- Capability is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Capability is at Learning (3 of 5), which means that leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help.
This family's technologies, ring by ring
- Adopt ring, 4 technologies. These are proven in operation, but your organisation is not yet ready to adopt them.
- Pilot ring, 2 technologies. These are real enough to test, and your organisation is ready to pilot them.
- Scan ring, 1 technology. It is at an early stage, so there is nothing to be ready for yet.
For Programme & Mission Delivery, overall readiness is 2.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Emerging (2 of 5).
Your organisation is not yet ready for any of this family's 7 technologies, all of which are in the Pilot or Adopt rings.
Customers now ask about AI in bids and require it in deliverables, and the contracts constrain what may be used. Programme teams are willing; they have no approved tools and no analysts.
- Its weakest dimensions are Regulatory and Capability, both scored at Emerging (2 of 5).
- Raising them one level, to Learning (3 of 5), would make your organisation ready to pilot these technologies.
- Raising all three dimensions to Developing (4 of 5) would make your organisation ready to adopt them.
- At Learning, Regulatory looks like this: a route to approval has been used in some areas, led by specialists, and some coming rule changes are known.
- Regulatory is scored at Emerging (2 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Regulatory is at Emerging (2 of 5), which means that the rules have been checked, and approval is possible case by case, with no set route.
- Acceptance is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Acceptance is at Learning (3 of 5), which means that in some parts of the organisation, those affected are consulted and their concerns acted on, led by specialists.
- Capability is scored at Emerging (2 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Capability is at Emerging (2 of 5), which means that a specialist team has looked at these technologies, but they are not a leadership priority and have no budget.
This family's technologies, ring by ring
- Adopt ring, 6 technologies. These are proven in operation, but your organisation is not yet ready to adopt them.
- Pilot ring, 1 technology. It is real enough to test, but your organisation is not yet ready to pilot it.
For Supply Chain & Partners, overall readiness is 2.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Emerging (2 of 5).
Your organisation is not yet ready for any of this family's 5 technologies in the Pilot or Adopt rings.
Cyber flow-down to suppliers is done and understood. AI flow-down, supplier AI assessment and data sharing with primes have not started, and the legal forms barely exist.
- Its weakest dimension is Capability, scored at Emerging (2 of 5).
- Raising Capability one level, to Learning (3 of 5), would make your organisation ready to pilot these technologies.
- Raising all three dimensions to Developing (4 of 5) would make your organisation ready to adopt them.
- At Learning, Capability looks like this: leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help.
- Regulatory is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Regulatory is at Learning (3 of 5), which means that a route to approval has been used in some areas, led by specialists, and some coming rule changes are known.
- Acceptance is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Acceptance is at Learning (3 of 5), which means that in some parts of the organisation, those affected are consulted and their concerns acted on, led by specialists.
- Capability is scored at Emerging (2 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Capability is at Emerging (2 of 5), which means that a specialist team has looked at these technologies, but they are not a leadership priority and have no budget.
This family's technologies, ring by ring
- Adopt ring, 3 technologies. These are proven in operation, but your organisation is not yet ready to adopt them.
- Pilot ring, 2 technologies. These are real enough to test, but your organisation is not yet ready to pilot them.
- Scan ring, 2 technologies. These are at an early stage, so there is nothing to be ready for yet.
For Workforce & Culture, overall readiness is 1.7 of 5, the average of the three dimensions; your organisation scored weakest on Acceptance and Capability, both at Beginning (1 of 5).
Your organisation is not yet ready for any of this family's 7 technologies, all of which are in the Pilot or Adopt rings.
The weakest domain. No literacy programme, nine cleared data scientists and three resignations this year, two unionised plants and no conversation started. Acceptance cannot be assumed.
- Its weakest dimensions are Acceptance and Capability, both scored at Beginning (1 of 5).
- Raising them two levels, to Learning (3 of 5), would make your organisation ready to pilot these technologies.
- Raising all three dimensions to Developing (4 of 5) would make your organisation ready to adopt them.
- At Emerging, Acceptance looks like this: views are known for a few high-profile uses, but there is no plan to answer concerns.
- Regulatory is scored at Learning (3 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Regulatory is at Learning (3 of 5), which means that a route to approval has been used in some areas, led by specialists, and some coming rule changes are known.
- Acceptance is scored at Beginning (1 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Acceptance is at Beginning (1 of 5), which means that those affected have not been asked, and openness goes no further than the law requires.
- Capability is scored at Beginning (1 of 5); starting a pilot needs it at Learning (3 of 5), and adopting needs it at Developing (4 of 5). Capability is at Beginning (1 of 5), which means that no one owns these technologies, and the skills, processes and systems they need are not in place.
This family's technologies, ring by ring
- Adopt ring, 5 technologies. These are proven in operation, but your organisation is not yet ready to adopt them.
- Pilot ring, 2 technologies. These are real enough to test, but your organisation is not yet ready to pilot them.
- An enterprise AI strategy tied to the business plan Leadership & Strategy
A written strategy that names the business outcomes AI is for, the money behind it and the measures it is judged by. Canada's defence department and the US Department of War have both published theirs, so a supplier without one is answering a customer that has.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Tamarack has an AI section in its annual plan and no strategy. The customer now asks for one in capability conversations; the gap is a leadership decision, not a technology one. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Leadership & Strategy, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: DND/CAF Artificial Intelligence Strategy (2024)
Evidence links
- DND/CAF AI Strategy (opens in a new tab) canada.ca
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed October 13, 2026.
- A named executive accountable for AI Leadership & Strategy
One executive owns AI across the firm, with a budget and a mandate, as the US Department of War did in creating its Chief Digital and AI Office, and as DND's strategy does in giving AI one home.
Maturity 9 of 9, Proven in service, in the Adopt ring. Actual system proven in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 5 of 5, Codified for replication.
What it changes: Four vice-presidents each own a piece. Nobody can say no to a pilot or yes to a platform. The cheapest move on the radar. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Leadership & Strategy, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is proven in service and your competitors most likely already run it, so the decision to adopt it is overdue. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: US Department of War, Chief Digital and AI Office, Responsible AI Toolkit
Evidence links
- CDAO RAI Toolkit (opens in a new tab) rai.tradewindai.com
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed October 13, 2026.
- An AI investment portfolio with stage gates Leadership & Strategy
AI spend run as a portfolio: every initiative staged from idea to scaled use, with a decision date and a kill criterion at each gate, and the whole reviewed by the executive together.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Eleven pilots are running on discretionary budgets. None has a decision date. A portfolio would end half and fund two properly. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Leadership & Strategy, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
- UK Defence AI Strategy (opens in a new tab) gov.uk
Last reviewed October 1, 2026.
- Scenario planning for AI-enabled competitors and customers Leadership & Strategy
Structured scenarios of how allies, adversaries, primes and new entrants will use AI over five years, used to test the product roadmap and the firm's position in the supply chain.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A method. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: The strategy assumes the customer buys what it bought last time. NATO's revised strategy commits the allies to adopting AI faster, so it will not. Worth a facilitated exercise before the next plan. Impact 4 of 5, Structural.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. Capability is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. For Leadership & Strategy, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is real enough to test. Your organisation is not yet ready to pilot it: Acceptance and Capability need to reach Learning (3 of 5) first. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: NATO Artificial Intelligence Strategy and Principles of Responsible Use (revised 2024)
Evidence links
- NATO AI Strategy (opens in a new tab) nato.int
- Our North, Strong and Free (opens in a new tab) canada.ca
Last reviewed September 17, 2026.
- A board-level AI risk appetite statement Leadership & Strategy
The board says in writing what AI risks the firm will and will not run: in products, in bids, in the back office and in what it shares with partners. The customer's own strategy says as much about its AI, in public.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A method. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: The board has discussed AI twice and decided nothing. A one-page appetite statement would unblock the governance board and the compliance team at once. Impact 3 of 5, Significant.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. Capability is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. For Leadership & Strategy, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is real enough to test. Your organisation is not yet ready to pilot it: Acceptance and Capability need to reach Learning (3 of 5) first. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: DND/CAF Artificial Intelligence Strategy (2024)
Evidence links
- DND/CAF AI Strategy (opens in a new tab) canada.ca
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
Last reviewed September 17, 2026.
- AI positioning in bids and capture strategy Leadership & Strategy
AI capability stated deliberately in bids and capture plans, because Canada's 2024 defence policy update and the department's AI strategy make AI a stated priority, so bids are read for it.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Bid teams add an AI paragraph when asked. Competitors lead with it. The capture process needs an AI thread, owned, from the first customer call. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Leadership & Strategy, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Our North, Strong and Free (Canada's 2024 defence policy update)
Evidence links
- Our North, Strong and Free (opens in a new tab) canada.ca
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed October 6, 2026.
- Partnering with defence innovation programmes Leadership & Strategy
Taking part in IDEaS, NATO DIANA and allied accelerators to co-fund AI work, meet the customer's problem owners and see what is coming before a request for proposal does.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Two IDEaS challenges entered, one won, neither followed into a product. The relationships exist; the pipeline from them does not. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Leadership & Strategy, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: IDEaS, Innovation for Defence Excellence and Security (DND)
Evidence links
- IDEaS programme (opens in a new tab) canada.ca
- NATO DIANA (opens in a new tab) diana.nato.int
Last reviewed October 6, 2026.
- Responsible AI principles adopted and published Governance, Ethics & Assurance
A short set of principles for how the firm builds and uses AI, aligned to NATO's principles of responsible use and to DND's own guiding principles for AI, published where customers and staff can read them.
Maturity 9 of 9, Proven in service, in the Adopt ring. Actual system proven in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 5 of 5, Codified for replication.
What it changes: Drafted in 2025 and never approved. Alignment to NATO's six principles and DND's guiding principles is a day's work; approval is the obstacle. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Governance, Ethics & Assurance, overall readiness is 3.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Learning (3 of 5). At Developing, Capability looks like this: senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary. This technology is proven in service and your competitors most likely already run it, so the decision to adopt it is overdue, and your organisation is nearly ready to adopt it: only Capability needs to reach Developing (4 of 5). The work now is integration, sequencing and deciding who moves first.
Where to verify: NATO Artificial Intelligence Strategy and Principles of Responsible Use (revised 2024)
Evidence links
- NATO AI Strategy (opens in a new tab) nato.int
- DND/CAF AI Guiding Principles (opens in a new tab) canada.ca
Last reviewed October 13, 2026.
- An AI governance board with real decision rights Governance, Ethics & Assurance
A standing body that approves AI uses, owns the register and the impact assessments, and can stop a deployment. DND's strategy gives AI one home and the Department of War's pathway makes such a body the centre of its programme.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Tamarack's AI working group advises and meets quarterly. Give it decision rights, a chair with authority and a monthly cadence and most of this domain moves. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Governance, Ethics & Assurance, overall readiness is 3.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Learning (3 of 5). At Developing, Capability looks like this: senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary. This technology is qualified and your peers are moving, so adopting it now keeps pace, and your organisation is nearly ready to adopt it: only Capability needs to reach Developing (4 of 5). The work now is integration, sequencing and deciding who moves first.
Where to verify: Responsible AI Strategy and Implementation Pathway (US Department of War, then Defense, 2022)
Evidence links
- DoW Responsible AI Pathway (opens in a new tab) media.defense.gov
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed October 13, 2026.
- Algorithmic impact assessments before deployment Governance, Ethics & Assurance
A structured assessment of each AI use before it goes live: who it affects, what could go wrong, what oversight it needs. Mandatory for the federal government's own automated decisions under the Directive on Automated Decision-Making, and increasingly asked of suppliers.
Maturity 9 of 9, Proven in service, in the Adopt ring. Actual system proven in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 5 of 5, Codified for replication.
What it changes: Not done for any of the eleven pilots. The federal template is free and fits a supplier with light changes; the compliance team could run it tomorrow. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Governance, Ethics & Assurance, overall readiness is 3.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Learning (3 of 5). At Developing, Capability looks like this: senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary. This technology is proven in service and your competitors most likely already run it, so the decision to adopt it is overdue, and your organisation is nearly ready to adopt it: only Capability needs to reach Developing (4 of 5). The work now is integration, sequencing and deciding who moves first.
Where to verify: Directive on Automated Decision-Making (Treasury Board of Canada)
Evidence links
- Directive on Automated Decision-Making (opens in a new tab) tbs-sct.canada.ca
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
Last reviewed October 20, 2026.
- A model and dataset inventory (an AI register) Governance, Ethics & Assurance
One list of every model, dataset and AI service in use, who owns it, what it touches and what it was assessed against. The register is what every audit asks for first.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: No one can list the firm's AI uses. A spreadsheet started this quarter would be the single most useful artefact in a customer audit. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Governance, Ethics & Assurance, overall readiness is 3.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Learning (3 of 5). At Developing, Capability looks like this: senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary. This technology is qualified and your peers are moving, so adopting it now keeps pace, and your organisation is nearly ready to adopt it: only Capability needs to reach Developing (4 of 5). The work now is integration, sequencing and deciding who moves first.
Where to verify: Guide on the use of generative AI (Treasury Board of Canada Secretariat)
Evidence links
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed October 1, 2026.
- Test, evaluation, verification and validation for AI components Governance, Ethics & Assurance
Disciplined testing of AI-bearing components against stated performance, robustness and safety criteria, as Directive 3000.09 demands of autonomous functions, as the Department of War's pathway builds into the life of every AI product, and as DND's new AI Centre exists to do.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The naval sensor team tests models well; the rest of the firm does not test them at all. A firm-wide standard, drawn from the naval practice, is the product differentiator in this domain. Impact 5 of 5, Defining.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Governance, Ethics & Assurance, overall readiness is 3.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Learning (3 of 5). At Developing, Capability looks like this: senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary. This technology is proven in operation but few run it yet, so adopting it now would lead, and your organisation is nearly ready to adopt it: only Capability needs to reach Developing (4 of 5). The work now is integration, sequencing and deciding who moves first.
Where to verify: Responsible AI Strategy and Implementation Pathway (US Department of War, then Defense, 2022)
Evidence links
- DoW Responsible AI Pathway (opens in a new tab) media.defense.gov
- Directive 3000.09 (opens in a new tab) esd.whs.mil
Last reviewed October 20, 2026.
- Human-in-the-loop design rules for autonomy Governance, Ethics & Assurance
Explicit rules for which decisions a person must make, confirm or be able to override, designed into the product rather than described in the manual. DND's strategy keeps a person in the loop wherever lethal force is involved, and the US directive sets the floor for the firm's US work.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Each product team decides for itself. A firm standard, mapped to DND's strategy, Directive 3000.09 and NATO's principles, would be reused in every bid. Impact 5 of 5, Defining.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Governance, Ethics & Assurance, overall readiness is 3.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Learning (3 of 5). At Developing, Capability looks like this: senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary. This technology is qualified and your peers are moving, so adopting it now keeps pace, and your organisation is nearly ready to adopt it: only Capability needs to reach Developing (4 of 5). The work now is integration, sequencing and deciding who moves first.
Where to verify: Directive 3000.09, Autonomy in Weapon Systems (US Department of War, 2023; issued as DoD Directive 3000.09)
Evidence links
- Directive 3000.09 (opens in a new tab) esd.whs.mil
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed October 20, 2026.
- Independent AI assurance and audit Governance, Ethics & Assurance
Third-party or internal-audit review of AI systems against the firm's principles and the customer's requirements. The UK and US defence ministries are building the assurance regimes suppliers will be audited under.
Maturity 3 of 9, Proof of concept, in the Scan ring. Analytical and experimental proof of concept. Technologies in the Scan ring have thin evidence and long timelines. A method. Low confidence in this placement; evidence level 2 of 5, Early data, no attribution.
What it changes: Nothing yet, and the market for assurers is immature: the 2025 review moved this back a level as early offerings fell away. Watch it, and ask internal audit to add AI to the next plan. Impact 4 of 5, Structural.
Regulatory is scored at Developing (4 of 5). Acceptance is scored at Developing (4 of 5). Capability is scored at Learning (3 of 5). For Governance, Ethics & Assurance, overall readiness is 3.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Learning (3 of 5). At Developing, Capability looks like this: senior leaders sponsor and fund these technologies, and teams everywhere use them with little specialist help, though results vary. This technology is one to watch, and watching it needs no readiness from your organisation. Assign someone to follow it, and do not budget against it; it enters the Pilot ring at TRL 4, once it has been validated in a laboratory.
Where to verify: UK Ministry of Defence, Defence Artificial Intelligence Strategy (2022)
Evidence links
- UK Defence AI Strategy (opens in a new tab) gov.uk
- DoW Responsible AI Pathway (opens in a new tab) media.defense.gov
Last reviewed September 24, 2026.
- Controlled-goods and ITAR screening of AI tools and data flows Security, Compliance & Export Control
Every AI tool, cloud service and dataset checked against the Controlled Goods Program and ITAR before controlled technical data touches it, with the same rigour applied to a chatbot as to a drawing.
Maturity 9 of 9, Proven in service, in the Adopt ring. Actual system proven in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 5 of 5, Codified for replication.
What it changes: Tamarack's export-control function is the strongest in the firm and has already blocked two public AI services. The firm is ready; the policy should say what is allowed, not only what is not. Impact 5 of 5, Defining.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Mastering (5 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. For Security, Compliance & Export Control, overall readiness is 4.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Developing (4 of 5). At Mastering, Regulatory looks like this: approval is routine and the same everywhere, and each rule change in view has a plan. This technology is proven in service and your competitors most likely already run it, so the decision to adopt it is overdue, and your organisation is ready to adopt it: every dimension is at Developing (4 of 5) or above. The work now is integration, sequencing and deciding who moves first.
Where to verify: Controlled Goods Program (Public Services and Procurement Canada)
Evidence links
- Controlled Goods Program (opens in a new tab) canada.ca
- Directorate of Defense Trade Controls (opens in a new tab) pmddtc.state.gov
Last reviewed October 20, 2026.
- Cyber certification controls, CPCSC and CMMC, across AI development environments Security, Compliance & Export Control
The cyber controls defence contracts now require on both sides of the border, Canada's CPCSC for National Defence work and CMMC for US work, built on the same controls; applied to the environments where models are built and data is prepared, not only to the engineering network.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The engineering network holds CPCSC Level 1 and is preparing for the Level 2 assessment Canada adds to contracts from 2027. The data science sandbox is outside the boundary; bring it in before the assessor does. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Mastering (5 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. For Security, Compliance & Export Control, overall readiness is 4.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Developing (4 of 5). At Mastering, Regulatory looks like this: approval is routine and the same everywhere, and each rule change in view has a plan. This technology is qualified and your peers are moving, so adopting it now keeps pace, and your organisation is ready to adopt it: every dimension is at Developing (4 of 5) or above. The work now is integration, sequencing and deciding who moves first.
Where to verify: Canadian Program for Cyber Security Certification (Public Services and Procurement Canada)
Evidence links
- CPCSC (opens in a new tab) canada.ca
- CMMC 2.0 (opens in a new tab) dodcio.defense.gov
Last reviewed October 6, 2026.
- Secure deployment of AI systems Security, Compliance & Export Control
Hardening, monitoring and incident response for deployed AI, following the 2024 joint guidance from the NSA, the Canadian Centre for Cyber Security and Five Eyes partners.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The security operations team knows the guidance and has nothing deployed to apply it to. Ready and waiting on the rest of the firm. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Mastering (5 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. For Security, Compliance & Export Control, overall readiness is 4.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Developing (4 of 5). At Mastering, Regulatory looks like this: approval is routine and the same everywhere, and each rule change in view has a plan. This technology is proven in operation but few run it yet, so adopting it now would lead, and your organisation is ready to adopt it: every dimension is at Developing (4 of 5) or above. The work now is integration, sequencing and deciding who moves first.
Where to verify: Deploying AI Systems Securely (NSA, CCCS and Five Eyes partners, 2024)
Evidence links
- Deploying AI Systems Securely (opens in a new tab) media.defense.gov
- CCCS Generative AI guidance (opens in a new tab) cyber.gc.ca
Last reviewed October 6, 2026.
- Sovereign or air-gapped model hosting Security, Compliance & Export Control
Running language and vision models on infrastructure the firm controls, in Canada, disconnected where the data requires it, so controlled data never leaves the boundary.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The IT team has priced a hosted model inside the controlled network. Approval sits with the AI working group, which has no decision rights (see Governance). Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Mastering (5 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. For Security, Compliance & Export Control, overall readiness is 4.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Developing (4 of 5). At Mastering, Regulatory looks like this: approval is routine and the same everywhere, and each rule change in view has a plan. This technology is proven in operation but few run it yet, so adopting it now would lead, and your organisation is ready to adopt it: every dimension is at Developing (4 of 5) or above. The work now is integration, sequencing and deciding who moves first.
Where to verify: Canadian Centre for Cyber Security, Generative artificial intelligence guidance
Evidence links
- CCCS Generative AI guidance (opens in a new tab) cyber.gc.ca
- Controlled Goods Program (opens in a new tab) canada.ca
Last reviewed September 24, 2026.
- Adversarial machine learning defences Security, Compliance & Export Control
Protection against models being poisoned in training, fooled in operation or extracted by querying. Guidance exists; practice is young everywhere.
Maturity 4 of 9, Validated in lab, in the Pilot ring. Component validated in a laboratory environment. Technologies in the Pilot ring are real enough to test. A method. Medium confidence in this placement; evidence level 2 of 5, Early data, no attribution.
What it changes: No one owns it. The naval sensor team is the natural home; a watching brief with a named owner is enough this year. Impact 4 of 5, Structural.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Developing (4 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Mastering (5 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Developing (4 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. For Security, Compliance & Export Control, overall readiness is 4.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Developing (4 of 5). At Mastering, Regulatory looks like this: approval is routine and the same everywhere, and each rule change in view has a plan. This technology is real enough to test, and your organisation is ready to pilot it: every dimension is at Learning (3 of 5) or above. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: Deploying AI Systems Securely (NSA, CCCS and Five Eyes partners, 2024)
Evidence links
- Deploying AI Systems Securely (opens in a new tab) media.defense.gov
- CCCS Generative AI guidance (opens in a new tab) cyber.gc.ca
Last reviewed September 24, 2026.
- Generative AI acceptable-use policy and data-leakage controls Security, Compliance & Export Control
A policy saying which generative tools staff may use, for what and with what data, enforced by technical controls, modelled on the federal guide.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: In force since 2025 and well understood; say so to the customer. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Acceptance is scored at Mastering (5 of 5), and adopting needs it at Developing (4 of 5), so it is ready. Capability is scored at Developing (4 of 5), and adopting needs it at Developing (4 of 5), so it is ready. For Security, Compliance & Export Control, overall readiness is 4.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Developing (4 of 5). At Mastering, Regulatory looks like this: approval is routine and the same everywhere, and each rule change in view has a plan. This technology is qualified and your peers are moving, so adopting it now keeps pace, and your organisation is ready to adopt it: every dimension is at Developing (4 of 5) or above. The work now is integration, sequencing and deciding who moves first.
Where to verify: Guide on the use of generative AI (Treasury Board of Canada Secretariat)
Evidence links
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
- CCCS Generative AI guidance (opens in a new tab) cyber.gc.ca
Last reviewed October 13, 2026.
- AI supply-chain provenance for models and training data Security, Compliance & Export Control
Knowing where a model came from, what it was trained on and who has touched it, as the firm already knows for a part. Allied programmes are writing the requirements now.
Maturity 3 of 9, Proof of concept, in the Scan ring. Analytical and experimental proof of concept. Technologies in the Scan ring have thin evidence and long timelines. A method. Low confidence in this placement; evidence level 2 of 5, Early data, no attribution.
What it changes: Early everywhere. The firm's parts-provenance discipline is the asset to extend; nothing to buy yet. Impact 4 of 5, Structural.
Regulatory is scored at Developing (4 of 5). Acceptance is scored at Mastering (5 of 5). Capability is scored at Developing (4 of 5). For Security, Compliance & Export Control, overall readiness is 4.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Developing (4 of 5). At Mastering, Regulatory looks like this: approval is routine and the same everywhere, and each rule change in view has a plan. This technology is one to watch, and watching it needs no readiness from your organisation. Assign someone to follow it, and do not budget against it; it enters the Pilot ring at TRL 4, once it has been validated in a laboratory.
Where to verify: US Department of War, Chief Digital and AI Office, Responsible AI Toolkit
Evidence links
- CDAO RAI Toolkit (opens in a new tab) rai.tradewindai.com
- NATO AI Strategy (opens in a new tab) nato.int
Last reviewed September 17, 2026.
- A data governance programme with named data owners Data & Digital Foundations
Every significant dataset has an owner, a steward, a classification and a quality measure, run as a programme rather than a project. The customers' own strategies, DND's and the Department of War's, make this the foundation of everything else.
Maturity 9 of 9, Proven in service, in the Adopt ring. Actual system proven in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 5 of 5, Codified for replication.
What it changes: Owners exist for engineering data and nowhere else. Production, sustainment and HR data are unowned, so three of the eleven pilots stalled on data they could not get. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Data & Digital Foundations, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is proven in service and your competitors most likely already run it, so the decision to adopt it is overdue. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: DND/CAF Artificial Intelligence Strategy (2024)
Evidence links
- DND/CAF AI Strategy (opens in a new tab) canada.ca
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed October 1, 2026.
- An enterprise data platform spanning engineering, production and sustainment Data & Digital Foundations
One governed place where engineering, production and in-service data can be joined and served to analysts and models, with access control that respects classification.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Three estates, no bridge. Predictive maintenance (see Engineering) is impossible until fielded-system data and design data meet. The largest single investment on the radar. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Data & Digital Foundations, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
- UK Defence AI Strategy (opens in a new tab) gov.uk
Last reviewed October 1, 2026.
- Data separation that AI pipelines respect Data & Digital Foundations
Controlled, classified and open data kept apart by design, so a model trained on one cannot leak another, and the separation is provable to an auditor.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The separation exists for documents and not for data pipelines. Export control and IT need to design it together before the data platform is built, not after. Impact 5 of 5, Defining.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Data & Digital Foundations, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Controlled Goods Program (Public Services and Procurement Canada)
Evidence links
- Controlled Goods Program (opens in a new tab) canada.ca
- CPCSC (opens in a new tab) canada.ca
Last reviewed October 20, 2026.
- MLOps: versioned models, pipelines and monitoring Data & Digital Foundations
Engineering discipline for models: versioned, tested, deployed through a pipeline and monitored for drift, as software has been for a decade.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The naval sensor team does this by hand. A platform choice would let the rest of the firm inherit it. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Data & Digital Foundations, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
- Deploying AI Systems Securely (opens in a new tab) media.defense.gov
Last reviewed September 24, 2026.
- Synthetic data for sensor and training-set gaps Data & Digital Foundations
Generated data standing in for operational data that is rare, classified or dangerous to collect, used to train and test models. Allied defence strategies name it among the enablers they will fund.
Maturity 5 of 9, Validated in context, in the Pilot ring. Component validated in a relevant environment. Technologies in the Pilot ring are real enough to test. A system. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: Relevant to the sensor products and nothing else yet. A pilot with the customer's blessing, once the governance board can approve one. Impact 3 of 5, Significant.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. Acceptance is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. For Data & Digital Foundations, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is real enough to test. Your organisation is not yet ready to pilot it: Regulatory and Capability need to reach Learning (3 of 5) first. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: DND/CAF Artificial Intelligence Strategy (2024)
Evidence links
- DND/CAF AI Strategy (opens in a new tab) canada.ca
- UK Defence AI Strategy (opens in a new tab) gov.uk
Last reviewed September 17, 2026.
- Digital twins of fielded systems as AI data sources Data & Digital Foundations
A living model of each fielded system fed by its sensors and maintenance records, which AI can learn from and plan against.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A system. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: The sustainment contract already pays for the data; the twin would be the product. A roadmap item once the data platform exists. Impact 4 of 5, Structural.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. Acceptance is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. For Data & Digital Foundations, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is real enough to test. Your organisation is not yet ready to pilot it: Regulatory and Capability need to reach Learning (3 of 5) first. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: UK Ministry of Defence, Defence Artificial Intelligence Strategy (2022)
Evidence links
- UK Defence AI Strategy (opens in a new tab) gov.uk
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed September 17, 2026.
- Common data standards with primes and customers Data & Digital Foundations
Agreed formats and interfaces for the data that flows between the firm, its primes and the customer, so AI on either side can use it. NATO interoperability work sets the direction.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A method. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: Every programme negotiates its own. Joining the standards conversation costs a person's time and buys a seat at the table. Impact 4 of 5, Structural.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. Acceptance is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. For Data & Digital Foundations, overall readiness is 2.0 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Beginning (1 of 5). At Emerging, Capability looks like this: a specialist team has looked at these technologies, but they are not a leadership priority and have no budget. This technology is real enough to test. Your organisation is not yet ready to pilot it: Regulatory and Capability need to reach Learning (3 of 5) first. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: NATO Artificial Intelligence Strategy and Principles of Responsible Use (revised 2024)
Evidence links
- NATO AI Strategy (opens in a new tab) nato.int
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed September 17, 2026.
- AI-assisted requirements and systems-engineering analysis Engineering & Product
Language models that read requirements, find conflicts and gaps, and trace them to design and test, inside a controlled environment.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A system. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: Engineers want it; export control forbids the public tools. The sovereign-hosted model (see Security) unblocks this and five other practices. Impact 3 of 5, Significant.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. For Engineering & Product, overall readiness is 3.0 of 5, with all three dimensions scored at Learning (3 of 5). At Developing, Regulatory looks like this: approval follows a set process everywhere, and coming rule changes are planned for, though results still vary. This technology is real enough to test, and your organisation is ready to pilot it: every dimension is at Learning (3 of 5) or above. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
- UK Defence AI Strategy (opens in a new tab) gov.uk
Last reviewed October 6, 2026.
- AI code generation under secure development controls Engineering & Product
Coding assistants used inside a controlled environment, with review, provenance and licence checks, so the productivity gain does not become a security or export finding.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Prohibited today. The controls are known and the sovereign model (see Security) is where it would run; the decision is a governance one. Engineering attrition cites it in exit interviews. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Engineering & Product, overall readiness is 3.0 of 5, with all three dimensions scored at Learning (3 of 5). At Developing, Regulatory looks like this: approval follows a set process everywhere, and coming rule changes are planned for, though results still vary. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Canadian Centre for Cyber Security, Generative artificial intelligence guidance
Evidence links
- CCCS Generative AI guidance (opens in a new tab) cyber.gc.ca
- CPCSC (opens in a new tab) canada.ca
Last reviewed October 6, 2026.
- Machine-learning sensor fusion and target recognition in products Engineering & Product
Machine learning in the firm's own naval and electronic-warfare products, fusing sensors and classifying contacts. Proven in allied programmes and expected by the customer.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Tamarack's strongest technical team and its best story. The firm is ahead of its own governance here, which is the risk to manage. Impact 5 of 5, Defining.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Engineering & Product, overall readiness is 3.0 of 5, with all three dimensions scored at Learning (3 of 5). At Developing, Regulatory looks like this: approval follows a set process everywhere, and coming rule changes are planned for, though results still vary. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: DND/CAF Artificial Intelligence Strategy (2024)
Evidence links
Last reviewed October 20, 2026.
- Safety and certification cases for AI-bearing systems Engineering & Product
Arguing to a certifier that a system with learned behaviour is safe, with evidence. The methods are immature everywhere, and the US directive demands them anyway.
Maturity 3 of 9, Proof of concept, in the Scan ring. Analytical and experimental proof of concept. Technologies in the Scan ring have thin evidence and long timelines. A method. Medium confidence in this placement; evidence level 2 of 5, Early data, no attribution.
What it changes: The sensor products will need this at the next upgrade. Start the safety case now; it will take longer than the software. Impact 5 of 5, Defining.
Regulatory is scored at Learning (3 of 5). Acceptance is scored at Learning (3 of 5). Capability is scored at Learning (3 of 5). For Engineering & Product, overall readiness is 3.0 of 5, with all three dimensions scored at Learning (3 of 5). At Developing, Regulatory looks like this: approval follows a set process everywhere, and coming rule changes are planned for, though results still vary. This technology is one to watch, and watching it needs no readiness from your organisation. Assign someone to follow it, and do not budget against it; it enters the Pilot ring at TRL 4, once it has been validated in a laboratory.
Where to verify: Directive 3000.09, Autonomy in Weapon Systems (US Department of War, 2023; issued as DoD Directive 3000.09)
Evidence links
- Directive 3000.09 (opens in a new tab) esd.whs.mil
- UK Defence AI Strategy (opens in a new tab) gov.uk
Last reviewed October 20, 2026.
- Predictive maintenance for fielded systems Engineering & Product
Models that forecast failures in fielded systems from sensor and maintenance data, turning in-service support from scheduled to predicted. Mature in allied fleets.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: A structural change to sustainment, and blocked by data (see Data). The sustainment customer would co-fund it; nobody has asked. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Engineering & Product, overall readiness is 3.0 of 5, with all three dimensions scored at Learning (3 of 5). At Developing, Regulatory looks like this: approval follows a set process everywhere, and coming rule changes are planned for, though results still vary. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed October 1, 2026.
- Simulation and AI-driven synthetic test environments Engineering & Product
Testing products and training operators in simulated environments where AI generates scenarios and adversary behaviour. A UK MOD priority, and a field allied accelerators fund.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The training division sells simulators and uses none of this in them. A product opportunity hiding as an internal tool. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. For Engineering & Product, overall readiness is 3.0 of 5, with all three dimensions scored at Learning (3 of 5). At Developing, Regulatory looks like this: approval follows a set process everywhere, and coming rule changes are planned for, though results still vary. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: NATO DIANA, Defence Innovation Accelerator for the North Atlantic
Evidence links
- NATO DIANA (opens in a new tab) diana.nato.int
- UK Defence AI Strategy (opens in a new tab) gov.uk
Last reviewed September 24, 2026.
- Autonomy software for uncrewed systems Engineering & Product
Software that lets uncrewed surface, underwater and aerial systems navigate, sense and act with bounded independence, under the human-control rules DND's strategy and the US directive set.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A system. High confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: Two bids lost to firms with an autonomy stack. Partner or build is a board decision this year, not a technology one. Impact 5 of 5, Defining.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. For Engineering & Product, overall readiness is 3.0 of 5, with all three dimensions scored at Learning (3 of 5). At Developing, Regulatory looks like this: approval follows a set process everywhere, and coming rule changes are planned for, though results still vary. This technology is real enough to test, and your organisation is ready to pilot it: every dimension is at Learning (3 of 5) or above. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: Directive 3000.09, Autonomy in Weapon Systems (US Department of War, 2023; issued as DoD Directive 3000.09)
Evidence links
- Directive 3000.09 (opens in a new tab) esd.whs.mil
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed October 20, 2026.
- AI-assisted proposal and bid drafting Programme & Mission Delivery
Language models drafting, checking and reconciling proposal text against the request, inside a controlled environment, with people deciding what is claimed.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The bid team uses personal tools on personal devices, against policy, because the firm offers nothing. The sovereign model fixes the risk and the productivity gap at once. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Programme & Mission Delivery, overall readiness is 2.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Emerging (2 of 5). At Learning, Regulatory looks like this: a route to approval has been used in some areas, led by specialists, and some coming rule changes are known. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Guide on the use of generative AI (Treasury Board of Canada Secretariat)
Evidence links
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
- CCCS Generative AI guidance (opens in a new tab) cyber.gc.ca
Last reviewed October 13, 2026.
- Schedule and cost risk prediction on programmes Programme & Mission Delivery
Models trained on the firm's own programme history that flag schedule and cost risk early, feeding the programme review rather than replacing it.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Twenty years of programme data sits in three systems. Cleaning it is the work; the modelling is routine. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Programme & Mission Delivery, overall readiness is 2.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Emerging (2 of 5). At Learning, Regulatory looks like this: a route to approval has been used in some areas, led by specialists, and some coming rule changes are known. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed October 1, 2026.
- Contract and compliance clause analysis with language models Programme & Mission Delivery
Models reading contracts and flow-down clauses for AI, cyber and export obligations the firm must meet or pass on, for a person to confirm.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Legal has three people and four hundred live contracts. A bounded pilot with the sovereign model, once governance can approve one. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Programme & Mission Delivery, overall readiness is 2.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Emerging (2 of 5). At Learning, Regulatory looks like this: a route to approval has been used in some areas, led by specialists, and some coming rule changes are known. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Guide on the use of generative AI (Treasury Board of Canada Secretariat)
Evidence links
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
- CPCSC (opens in a new tab) canada.ca
Last reviewed September 24, 2026.
- AI in in-service support and sustainment planning Programme & Mission Delivery
Analytics and models planning spares, maintenance windows and technician deployment across the fleets the firm supports.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The sustainment business is 40 per cent of revenue and has no analyst. The customer measures availability; this is how the firm would move it. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Programme & Mission Delivery, overall readiness is 2.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Emerging (2 of 5). At Learning, Regulatory looks like this: a route to approval has been used in some areas, led by specialists, and some coming rule changes are known. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: DND/CAF Artificial Intelligence Strategy (2024)
Evidence links
- DND/CAF AI Strategy (opens in a new tab) canada.ca
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed October 1, 2026.
- Customer requirements for AI content in deliverables Programme & Mission Delivery
Knowing, tracing and evidencing what the customer requires of any AI inside a deliverable: documentation, test evidence, human-control design, data provenance.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A method. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: Requirements arrive in each contract and are met by each team alone. A firm-wide requirements library would cut bid risk and audit time. Impact 4 of 5, Structural.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. Acceptance is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. For Programme & Mission Delivery, overall readiness is 2.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Emerging (2 of 5). At Learning, Regulatory looks like this: a route to approval has been used in some areas, led by specialists, and some coming rule changes are known. This technology is real enough to test. Your organisation is not yet ready to pilot it: Regulatory and Capability need to reach Learning (3 of 5) first. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: Responsible AI Strategy and Implementation Pathway (US Department of War, then Defense, 2022)
Evidence links
- DoW Responsible AI Pathway (opens in a new tab) media.defense.gov
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed October 13, 2026.
- Decision-support analytics for programme reviews Programme & Mission Delivery
Dashboards and analytics that put cost, schedule, risk and quality in one view for programme reviews, with the data refreshed rather than retyped.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Monthly reviews run on slides built by hand over a week. Mature tooling; the obstacle is the three data estates. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Programme & Mission Delivery, overall readiness is 2.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Emerging (2 of 5). At Learning, Regulatory looks like this: a route to approval has been used in some areas, led by specialists, and some coming rule changes are known. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed September 24, 2026.
- Knowledge retrieval over programme records Programme & Mission Delivery
A model answering questions from the firm's own programme, engineering and lessons-learned records, with the source cited, inside the controlled boundary.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Forty years of records nobody can search. The highest-demand request from staff and a natural first use of the sovereign model. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Programme & Mission Delivery, overall readiness is 2.3 of 5, the average of the three dimensions; your organisation scored weakest on Regulatory and Capability, both at Emerging (2 of 5). At Learning, Regulatory looks like this: a route to approval has been used in some areas, led by specialists, and some coming rule changes are known. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Canadian Centre for Cyber Security, Generative artificial intelligence guidance
Evidence links
- CCCS Generative AI guidance (opens in a new tab) cyber.gc.ca
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
Last reviewed October 13, 2026.
- Supplier risk analytics Supply Chain & Partners
Models and data services that watch suppliers for financial, geopolitical, cyber and single-source risk, where Canada's 2024 policy update makes the resilience of the defence supply chain a national priority.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Supply chain tracks 1,100 suppliers in a spreadsheet. Commercial services exist; procurement has not been asked to buy one. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Supply Chain & Partners, overall readiness is 2.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Emerging (2 of 5). At Learning, Capability looks like this: leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Our North, Strong and Free (Canada's 2024 defence policy update)
Evidence links
- Our North, Strong and Free (opens in a new tab) canada.ca
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed October 6, 2026.
- Flowing AI governance requirements to suppliers Supply Chain & Partners
Passing the firm's AI, cyber and data obligations down to suppliers in contract, as CPCSC and CMMC already force for cyber, and checking they are met.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A method. High confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: Cyber flow-down is done and AI flow-down is not. The same contract clauses, extended; legal and procurement can do it this year. Impact 4 of 5, Structural.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. For Supply Chain & Partners, overall readiness is 2.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Emerging (2 of 5). At Learning, Capability looks like this: leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help. This technology is real enough to test, and your organisation is nearly ready to pilot it: only Capability needs to reach Learning (3 of 5). Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: Canadian Program for Cyber Security Certification (Public Services and Procurement Canada)
Evidence links
- CPCSC (opens in a new tab) canada.ca
- DoW Responsible AI Pathway (opens in a new tab) media.defense.gov
Last reviewed September 24, 2026.
- Demand forecasting and inventory optimisation for spares Supply Chain & Partners
Models forecasting spares demand across supported fleets and setting stock levels, a settled practice in commercial and defence logistics.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A system. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Running in one sustainment programme since 2024; spread it to the other four programmes. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Supply Chain & Partners, overall readiness is 2.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Emerging (2 of 5). At Learning, Capability looks like this: leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed October 1, 2026.
- Counterfeit and parts-provenance detection Supply Chain & Partners
Image and data analysis that spots counterfeit or misdescribed parts and traces provenance through the chain, which Canada's 2024 policy update makes a priority for the defence industrial base.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A system. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: The quality team does this by eye and does it well. The 2025 review moved this up on a vendor's trial and back when the trial ended. A pilot with the receiving inspection team is bounded and defensible. Impact 4 of 5, Structural.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Capability is scored at Emerging (2 of 5), and a pilot needs it at Learning (3 of 5), one level up. For Supply Chain & Partners, overall readiness is 2.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Emerging (2 of 5). At Learning, Capability looks like this: leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help. This technology is real enough to test, and your organisation is nearly ready to pilot it: only Capability needs to reach Learning (3 of 5). Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: Our North, Strong and Free (Canada's 2024 defence policy update)
Evidence links
- Our North, Strong and Free (opens in a new tab) canada.ca
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed September 17, 2026.
- Shared AI test data and models with primes under controlled-goods terms Supply Chain & Partners
Agreements and mechanisms for exchanging training data and models with primes and partners without breaching controlled-goods and ITAR rules. The legal forms barely exist.
Maturity 3 of 9, Proof of concept, in the Scan ring. Analytical and experimental proof of concept. Technologies in the Scan ring have thin evidence and long timelines. A method. Medium confidence in this placement; evidence level 2 of 5, Early data, no attribution.
What it changes: Two primes have asked. Export control says no because nobody has designed a yes. A legal and policy question for the year, not a purchase. Impact 4 of 5, Structural.
Regulatory is scored at Learning (3 of 5). Acceptance is scored at Learning (3 of 5). Capability is scored at Emerging (2 of 5). For Supply Chain & Partners, overall readiness is 2.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Emerging (2 of 5). At Learning, Capability looks like this: leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help. This technology is one to watch, and watching it needs no readiness from your organisation. Assign someone to follow it, and do not budget against it; it enters the Pilot ring at TRL 4, once it has been validated in a laboratory.
Where to verify: Controlled Goods Program (Public Services and Procurement Canada)
Evidence links
- Controlled Goods Program (opens in a new tab) canada.ca
- Directorate of Defense Trade Controls (opens in a new tab) pmddtc.state.gov
Last reviewed September 17, 2026.
- AI capability assessment in supplier selection Supply Chain & Partners
Asking suppliers how they use and govern AI as part of selection and review, as the firm's own customers are beginning to ask it.
Maturity 3 of 9, Proof of concept, in the Scan ring. Analytical and experimental proof of concept. Technologies in the Scan ring have thin evidence and long timelines. A method. Low confidence in this placement; evidence level 2 of 5, Early data, no attribution.
What it changes: Not asked today. Add three questions to the supplier questionnaire; the answers will surprise. Impact 3 of 5, Significant.
Regulatory is scored at Learning (3 of 5). Acceptance is scored at Learning (3 of 5). Capability is scored at Emerging (2 of 5). For Supply Chain & Partners, overall readiness is 2.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Emerging (2 of 5). At Learning, Capability looks like this: leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help. This technology is one to watch, and watching it needs no readiness from your organisation. Assign someone to follow it, and do not budget against it; it enters the Pilot ring at TRL 4, once it has been validated in a laboratory.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
- UK Defence AI Strategy (opens in a new tab) gov.uk
Last reviewed September 17, 2026.
- Collaboration with allied defence innovation accelerators Supply Chain & Partners
Working with DIANA, IDEaS and allied accelerators as a route to partners, funding and the customer's problem set, rather than as a sponsorship.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: One DIANA challenge watched, none entered. The Leadership domain's partnering practice is the same gap seen from the supply side. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Capability is scored at Emerging (2 of 5), and adopting needs it at Developing (4 of 5), two levels up. For Supply Chain & Partners, overall readiness is 2.7 of 5, the average of the three dimensions; your organisation scored weakest on Capability, at Emerging (2 of 5). At Learning, Capability looks like this: leaders see the value of these technologies and back a pilot, but non-specialists still need specialist help. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: NATO DIANA, Defence Innovation Accelerator for the North Atlantic
Evidence links
- NATO DIANA (opens in a new tab) diana.nato.int
- IDEaS programme (opens in a new tab) canada.ca
Last reviewed October 6, 2026.
- AI literacy for every employee Workforce & Culture
Everyone understands what AI is, what the firm permits, how to use it well and when to distrust it. Public services are rolling it out across their workforces.
Maturity 9 of 9, Proven in service, in the Adopt ring. Actual system proven in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 5 of 5, Codified for replication.
What it changes: Nothing offered beyond the acceptable-use policy. A half-day course for 2,500 people is cheap and is where acceptance starts. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Workforce & Culture, overall readiness is 1.7 of 5, the average of the three dimensions; your organisation scored weakest on Acceptance and Capability, both at Beginning (1 of 5). At Emerging, Acceptance looks like this: views are known for a few high-profile uses, but there is no plan to answer concerns. This technology is proven in service and your competitors most likely already run it, so the decision to adopt it is overdue. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Guide on the use of generative AI (Treasury Board of Canada Secretariat)
Evidence links
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
Last reviewed October 13, 2026.
- A security-cleared AI and data talent pipeline Workforce & Culture
Hiring, clearing, developing and keeping people who can build AI inside a controlled environment. Allied defence strategies name talent among their hardest constraints.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A method. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: Fourteen data scientists, nine cleared, three resignations this year. Clearance through the Contract Security Program is the bottleneck; co-op pipelines with cleared universities are the fix other Canadian firms use. Impact 5 of 5, Defining.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. Capability is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. For Workforce & Culture, overall readiness is 1.7 of 5, the average of the three dimensions; your organisation scored weakest on Acceptance and Capability, both at Beginning (1 of 5). At Emerging, Acceptance looks like this: views are known for a few high-profile uses, but there is no plan to answer concerns. This technology is real enough to test. Your organisation is not yet ready to pilot it: Acceptance and Capability need to reach Learning (3 of 5) first. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: DND/CAF Artificial Intelligence Strategy (2024)
Evidence links
- DND/CAF AI Strategy (opens in a new tab) canada.ca
- UK Defence AI Strategy (opens in a new tab) gov.uk
Last reviewed October 13, 2026.
- Role-based AI training for engineers and programme managers Workforce & Culture
Training by role: engineers in building and testing models, programme managers in buying and governing them, leaders in deciding about them.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Engineers learn on their own time. DND's strategy makes talent and training a line of effort and the Department of War's pathway lays out the curriculum by role; borrow both. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Workforce & Culture, overall readiness is 1.7 of 5, the average of the three dimensions; your organisation scored weakest on Acceptance and Capability, both at Beginning (1 of 5). At Emerging, Acceptance looks like this: views are known for a few high-profile uses, but there is no plan to answer concerns. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Responsible AI Strategy and Implementation Pathway (US Department of War, then Defense, 2022)
Evidence links
- DoW Responsible AI Pathway (opens in a new tab) media.defense.gov
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed October 1, 2026.
- Change management for AI-altered roles Workforce & Culture
Deliberate redesign of jobs that AI changes, with the people in them, before the tools arrive rather than after.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: Not started. The bid team and the quality inspectors are the first two roles that will change; begin with them. Impact 4 of 5, Structural.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Workforce & Culture, overall readiness is 1.7 of 5, the average of the three dimensions; your organisation scored weakest on Acceptance and Capability, both at Beginning (1 of 5). At Emerging, Acceptance looks like this: views are known for a few high-profile uses, but there is no plan to answer concerns. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: UK Ministry of Defence, Defence Artificial Intelligence Strategy (2022)
Evidence links
- UK Defence AI Strategy (opens in a new tab) gov.uk
- DND/CAF AI Strategy (opens in a new tab) canada.ca
Last reviewed September 24, 2026.
- Communities of practice and internal AI champions Workforce & Culture
Volunteer networks that spread what works, surface problems early and give the governance board eyes across the firm.
Maturity 8 of 9, Qualified, in the Adopt ring. System complete and qualified through test and demonstration. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: An unofficial channel of about sixty people exists. Recognise it, fund its time and give it a route to the governance board. Impact 2 of 5, Useful.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Workforce & Culture, overall readiness is 1.7 of 5, the average of the three dimensions; your organisation scored weakest on Acceptance and Capability, both at Beginning (1 of 5). At Emerging, Acceptance looks like this: views are known for a few high-profile uses, but there is no plan to answer concerns. This technology is qualified and your peers are moving, so adopting it now keeps pace. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: Data, Analytics and AI Adoption Strategy (US Department of War, then Defense, 2023)
Evidence links
- DoW Data, Analytics and AI Adoption Strategy (opens in a new tab) media.defense.gov
- CDAO RAI Toolkit (opens in a new tab) rai.tradewindai.com
Last reviewed September 24, 2026.
- Union and workforce consultation on AI in the workplace Workforce & Culture
Consulting the unions and the workforce on where AI will be used and how people are protected, as Canadian public-sector practice expects and the firm's collective agreements will require.
Maturity 6 of 9, Demonstrated, in the Pilot ring. System demonstrated in a relevant environment. Technologies in the Pilot ring are real enough to test. A method. Medium confidence in this placement; evidence level 3 of 5, Causal evidence.
What it changes: Two of three plants are unionised and no conversation has started. Start it before the first tool lands, or the first tool becomes the conversation. Impact 3 of 5, Significant.
This technology is real enough to test, so the question is whether your organisation is ready to pilot it. Regulatory is scored at Learning (3 of 5), and a pilot needs it at Learning (3 of 5), so it is ready. Acceptance is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. Capability is scored at Beginning (1 of 5), and a pilot needs it at Learning (3 of 5), two levels up. For Workforce & Culture, overall readiness is 1.7 of 5, the average of the three dimensions; your organisation scored weakest on Acceptance and Capability, both at Beginning (1 of 5). At Emerging, Acceptance looks like this: views are known for a few high-profile uses, but there is no plan to answer concerns. This technology is real enough to test. Your organisation is not yet ready to pilot it: Acceptance and Capability need to reach Learning (3 of 5) first. Run a scoped pilot with a decision date and a kill criterion; the technology enters the Adopt ring at TRL 7, once an operational prototype has been demonstrated.
Where to verify: Guide on the use of generative AI (Treasury Board of Canada Secretariat)
Evidence links
- TBS Guide on the use of generative AI (opens in a new tab) canada.ca
- Voluntary Code of Conduct (opens in a new tab) ised-isde.canada.ca
Last reviewed September 17, 2026.
- Ethical escalation routes for AI concerns Workforce & Culture
A clear way for any employee to raise a concern about an AI use, inside the firm's existing ethics and whistleblowing routes, with a duty to answer.
Maturity 7 of 9, Operational prototype, in the Adopt ring. System prototype demonstrated in an operational environment. Technologies in the Adopt ring are proven in operation. A method. High confidence in this placement; evidence level 4 of 5, Replicated and independently evaluated.
What it changes: The ethics line exists and does not mention AI. A paragraph in the policy and a briefing to the ethics officer; a week's work. Impact 3 of 5, Significant.
This technology is proven in operation and already adopted elsewhere, so the question is whether your organisation is ready to adopt it. Regulatory is scored at Learning (3 of 5), and adopting needs it at Developing (4 of 5), one level up. Acceptance is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. Capability is scored at Beginning (1 of 5), and adopting needs it at Developing (4 of 5), three levels up. For Workforce & Culture, overall readiness is 1.7 of 5, the average of the three dimensions; your organisation scored weakest on Acceptance and Capability, both at Beginning (1 of 5). At Emerging, Acceptance looks like this: views are known for a few high-profile uses, but there is no plan to answer concerns. This technology is proven in operation but few run it yet, so adopting it now would lead. Your organisation is not yet ready to adopt it: all three dimensions need to reach Developing (4 of 5) first. The work now is integration, sequencing and deciding who moves first.
Where to verify: NATO Artificial Intelligence Strategy and Principles of Responsible Use (revised 2024)
Evidence links
- NATO AI Strategy (opens in a new tab) nato.int
- DoW Responsible AI Pathway (opens in a new tab) media.defense.gov
Last reviewed September 17, 2026.
Technology radar by Trailblazed Ventures