Line of Effort 1: Fielding and Employing AI Capabilities

The challenge

AI will be foundational to Defence modernization. The fusion of AI, networking, predictive and prescriptive analytics, ML, and robotics will be critical to equip DND/CAF and digitally transform defence. This will provide the capabilities needed to ensure operational superiority over our adversaries and technological parity and interoperability with our allies. The sheer volume, diversity and complexity of data produced by modern sensors has already overwhelmed the capacity of human analysts to process and interpret it. This problem will only increase as legacy systems are upgraded and replaced.

Despite this, we are only beginning to identify the AI-enabled capabilities we need and how to achieve them. Although DND/CAF possesses areas of AI expertise, its maturity, integration, and implementation are still at an early stage of development. Our current efforts are divided across environments and commands, with each element approaching AI enablement and capabilities separately, and hampered by data quality issues. Effectively fielding and employing AI capabilities will require a concerted, collaborative effort until DND/CAF is more digitally advanced and technologically proficient.

AI is beginning to proliferate into many facets of operations and the business of Defence. DND/CAF needs to be fully aware of the benefits and challenges of AI to aid in the responsible, transparent use of this powerful tool while minimizing the potential risks. Any AI that is employed by the CAF on operations, including AI and AI-enabled information obtained from allies or partners, will likely receive scrutiny and this requires additional attention by operational authorities to ensure they are aligned with Canadian law and policies.

The accountability and responsibility of AI systems may crosscut existing structures within and outside DND/CAF. Further, because of their application to national defence, some DND/CAF use cases may fall outside the policy provided by the Treasury Board of Canada Secretariat, and the gap between the development of AI and legislative and policy coverage will widen further. Where laws and policies do not apply, or have not yet been developed, L1s must work collaboratively to develop appropriate guidance to ensure the responsible and transparent use of AI. We must balance testing, validation and risk mitigation with the agility required to adopt AI technology at the speed of relevance.

What we must do

We must responsibly and transparently use AI to provide capabilities for operations and to solve key business problems. We must quickly build the internal expertise and experience to assess, advise, field, and manage AI tools and applications for operations and the business of defence. We must also provide collaborative mechanisms to support the use of AI during operations, whether the technology originates within DND/CAF or is shared with us by our allies or partners.

We must align our investments in AI with existing priorities. Fielding and employing AI capabilities will not happen all at once. DND/CAF will therefore prioritize the fielding of AI systems that support commitments within the Departmental Plan to augment decision-making and processes, and to improve the capabilities and performance of our personnel by freeing them from repetitive or dangerous tasks. Initially, efforts will be focused on enabling CAF operational capabilities as well as efforts that support the business of defence, particularly those that support CAF reconstitution and readiness. These capability areas can be directly linked to Departmental Plan priorities to overcome existing DND/CAF challenges. Although these efforts will need to be collaborative and horizontally executed at the lowest level, effective governance will be required to ensure they are coordinated, prioritized, and implemented within DND/CAF plans and priorities. Regular assessments will be required to monitor and report on progress.

We will identify or develop tools to support the rapid and safe design, procurement, integration, and scaling of AI. The initial focus will be on tools to support the evaluation of AI for operations but will also include problem definition tools to help identify when problems are amenable to an AI solution, guidelines to assist project teams and support staff in procuring them, and both quantitative and qualitative risk assessment tools to identify and mitigate risks to privacy, security, safety, human rights, and GBA Plus considerations.

How we will do this

  1. Establish an internal DND/CAF AI Centre (DCAIC). Like those of our defence partners (below) the DCAIC will act as a hub of AI expertise and an accelerator for experimentation, testing, evaluation, and fielding AI. This Centre will establish AI adoption processes for developing and onboarding AI technologies in collaboration with key stakeholders and provide support and advice to all to enable safe and responsible AI adoption and use.
  2. Conduct an AI maturity assessment and develop metrics for measuring our progress and performance. An assessment of current levels of AI implementation and capacity will establish a baseline against which the success of the AI Strategy and our own AI enablement can be measured using key performance indicators (KPI).
  3. Develop supporting mechanisms and tools to support AI growth. DND/CAF will need standards, guardrails, checklists and internal processes to enable the rapid and safe use, design, procurement, integration, risk management and scaling of AI. The DCAIC will lead this effort in collaboration with all stakeholders.
  4. Develop governance for AI to align resources and goals. Options for effective governance of AI will be developed and considered as part of the overall governance structure within DND/CAF. This governance will ensure that efforts remain aligned to existing priorities, policy and capacity.

A DND/CAF AI Centre (DCAIC)

Within the last five years, three of Canada’s closest defence partners have established centres of excellence to accelerate AI experimentation and scaling across the Defence enterprise. In 2018, the U.S. Department of Defense (DoD) established the Joint AI Center to accelerate AI development and deployment, and to support the implementation of the DoD AI Strategy. The Australian Defence AI Centre was established in the same year, and the U.K. Ministry of Defence launched its own Defence AI Centre in April 2022 to serve as an accelerator of concepts to operationalize and scale across defence, with key pillars focusing on robotics, digital, and responsible AI. In addition, AI accelerators for defence were endorsed as best practice by the U.S. National Security Commission on Artificial Intelligence in its final report in 2021.

To reach its AI goals, DND/CAF needs an AI Centre of its own. Working cross-functionally with teams currently exploiting or experimenting with AI, the Centre would serve as a hub of AI expertise for the Defence Team. It would identify and support the integration of capabilities developed by industry, academia, or allies, facilitate the creation of a common repository of AI applications, techniques, and data, develop common AI applications to meet shared challenges across the organization, and support local AI development within the commands and environments. It would also provide guidance on related issues required for successful AI implementation, such as policy, ethics, GBA Plus, procurement, and training. The DCAIC would work with stakeholders and users from the Warfare Centres, DRDC research centres, Joint Operations Fusion Lab, Innovation for Defence Excellence and Security (IDEaS), and other Defence Team branches (L1s), to coordinate the setting of organizational priorities for AI, ensure internal interoperability and economies of effort and scale, and help to ensure that successful initiatives were able to be appropriately scaled. It would also collaborate with Five Eyes partners and NATO allies on standards and guidelines for interoperability and responsible use. It would also have an outreach role with academia, industry and civil society, and with other government departments and review bodies. In this role, it would help foster the growth of the Canadian defence AI ecosystem, identify and promote the uptake of promising solutions within DND/CAF, and provide an environment in which defence and security challenges could be discussed.

Page details

Date modified: