Line of Effort 5: Partnerships

The challenge

The state of the art in AI is collaborative—and often outside defence. Internally, implementing and leveraging AI will take the collaboration of many within the Defence Team, however, that is only the beginning. Globally, although defence research once claimed the technological leading edge, the most advanced AI techniques and applications are now typically found in industry, academia, and the open-source community. Consequently, DND/CAF must be open to procure, co-design, co-develop, test and validate with trusted partners. Adopting AI-enabled technologies will also require international cooperation between nations sharing the same values to develop and agree upon principles, policies and standards to build and protect a secured digital infrastructure and supply chain with the right analytics.

At present, AI in DND/CAF needs the coordination that would come from partnerships. DND/CAF must continue to actively pursue opportunities for defence, safety and security collaboration with partners from other government departments and agencies, industry, academia, and international allies to maximize the strategic advantages of Canada’s innovation ecosystem and to secure it against adversaries and threats. Engagement activities must span the entire innovation spectrum, including scientific and technical information and personnel exchange, shared data, jointly developed frameworks, trials, experiments, advanced concept technology demonstrations, and consultation with communities of practice.

What we must do

We must ensure that innovation is adaptable and interoperable. Developing a strategic AI vision that aligns with Canada’s defence allies, including the Five Eyes partnership, the Technical Cooperation Program (TTCP), and our NATO Allies, will allow DND/CAF to prioritize and further develop AI research on interoperable capabilities, while sharing current best practice on building a military AI capability and educating personnel in its use. It will also afford opportunities to invest strategically in technologies that complement those of our defence partners.

We must work to build and leverage a defence and security AI ecosystem. Both Canada and DND/CAF benefit from the growth of talent and competition in AI. It is in DND/CAF’s interest to nurture a vibrant, diverse, agile, and responsive defence innovation ecosystem drawing on that already created by the Pan-Canadian AI Strategy and administered by the Canadian Institute for Advanced Research (CIFAR). The experience of CIFAR proves clearly that this can be done. However, this will require innovative tools for engagement with industry, academia, and non-traditional partners to maximise the identification and leveraging of joint capabilities, speed up procurement, and facilitate access to experts for short-term engagements. It will require that we identify dual use and defence-specific opportunities, and partner with small and medium enterprises to field limited-use and scalable capabilities. It will also require efforts to overcome the obstacles to engaging industry effectively, to aligning digital infrastructure that supports external collaboration, and to protecting intellectual property. Finally, it will require a long-term commitment to invest beyond experimentation to scale.

We must continue to innovate in partnership with external sectors. Strong partnerships with the private sector and academic institutions at the leading edge of AI advances will be vital at every stage in the AI life cycle, from research, validation, certification, deployment, sustainment, and decommissioning. DND/CAF must increase partnerships with academia and industry to identify, fund, and enable breakthroughs in AI. This collaboration should include traditional funding models and existing initiatives such as Mobilizing Insights in Defence and Security (MINDS) and Innovation for Defence Excellence and Security (IDEaS), but also increased use of innovative approaches such as grand challenges, hackathons, and creative design sessions with start-ups and integrators to co-design and co-develop AI-enabled solutions at the leading edge of the technology. We must also explore non-traditional partnerships. Defence has much to learn from non-traditional partners such as civil society, non-technical branches of academia, and the open-source community.

We must deepen collaboration with other government departments to yield efficiencies and a common approach. For economy of effort and interoperability, DND/CAF must work towards co-developing and sharing AI capabilities with other government departments, particularly with those departments and agencies with whom we share responsibilities for national security. A whole-of-government approach will enable transfer of technology, better external relationship-building and partnerships, and the acquisition of AI technology that can benefit all areas.

We must establish guidelines to ensure that our partnerships are trusted and secure. We know that our allies, industry and academia are constantly targeted by potential adversaries, making our reliance on AI partners a potential source of strategic vulnerability. This is particularly true for dual-use technologies. Academic researchers often work in global networks whose boundaries and membership can be hard to secure, while Canadian companies continue to be subject to industrial espionage and can experience supply chain vulnerabilities. We must therefore invest in research and the development of guidelines to ensure that our partnerships are appropriately secured against external threats.

How we will do this

1. Reinforce strategic partnerships and cooperation with allies on new capabilities, best practices, and lesson learned. This should include furthering our strategic collaboration with the Five Eyes, TTCP, and NATO, but also deepening bilateral cooperation with trusted nations on specific areas of common interest.
2. Contribute to improving the procurement process to support development and acquisition of AI that balances security, safety, and ethics with the imperative to maintain flexibility and agility. We must continue to work with Public Services and Procurement Canada (PSPC) to identify the specific requirements and opportunities to simplify and accelerate procurement processes. This will lower administrative barriers for Canadian innovators to onboard, increase supplier diversity, and enhance economic and social opportunities for underrepresented groups. Procurement changes should also encourage organizations to identify AI and its technical foundations such as data and digital infrastructure within capability requirements.
3. Contribute to activities that encourage secure and reliable data infrastructures and sharing with partners and allies, working on common datasets to develop interoperable solutions.
4. Enhance connections with academia to encourage skill development in support of defence requirements. We must encourage innovative partnerships through personnel exchanges, such as the Mobility initiative to enable personnel exchange between IDEaS partners. We must also look for opportunities for new training paths to develop a talent pipeline for defence.
5. Further develop and foster the innovation ecosystem for defence and security in partnership with the Pan-Canadian Innovation Ecosystem in AI. We must explore avenues for skill-sharing and assignments to deploy talent flexibly across sectors. We will also need to expand linkages and support for innovative solutions to defence challenges and foster processes to scale these initiatives across the Defence Team. This could include existing vehicles such as the DRDC research centres, and the MINDS and IDEaS programs, but should also extend to new funding lines to support priority military capabilities and mechanisms to provide challenges to industry. In doing so we must be strategic about research and development areas where we can build on Canada’s existing strengths.
6. Work with government partners to ensure defence and security interests are included in the design of the Canadian innovation ecosystem. The innovation ecosystem must be secured against adversarial actions and supported to build linkages to the defence environment. This should include attention to the security of the supply chain and the need for export controls on sensitive Canadian technology. We must also ensure cohesion and coherence in the regulatory space, reducing the administrative burdens of compliance for non-experts.