by Matthew Lombardi & Max Seltzer
March 23, 2018
In late 2017, Vladimir Putin declared that whichever country wins the artificial intelligence (AI) race will “rule the world.” Putin knows the power of low cost cyberwarfare – despite running an economy smaller than Canada’s, the Russian leader has outsized his geopolitical influence with targeted investments in comparatively old-fashioned digital technologies. Geopolitical risk expert Ian Bremmer has declared the global AI arms race one of the top global security threats of 2018, as China — another western rival fine-tuning its military apparatus — has set a public goal of matching the United States in AI by 2020.
Technology experts are sounding the alarm that if authoritarian regimes successfully leapfrog the U.S. and its allies in the AI race, the implications will reshape the western-dominated economic order that has persisted since the end of WWII. Such a development will also upend the geopolitical map, and spell the end of the western-backed security order that has existed since the formation of NATO.
Canada can tackle this issue by taking action to link the economic strength of our domestic AI ecosystem to our national security, and the solidification of a lasting western security advantage. We can do this by championing the creation of Sovereign AI Patent Funds amongst NATO allies, designed to spur defence investments in domestic AI research and development ecosystems.
Such an initiative should also qualify against NATO’s 2% of national GDP spending targets for defence – a threshold called for by the security alliance that most members, including Canada, do not currently meet.
What are the different AI development paradigms?
Canada is a critical player in shaping how the AI arms race unfolds. As a global hub for AI talent and development, we confront this security challenge from a position of economic strength, with world-leading AI hubs in Montreal and Toronto, backed by a philosophical commitment to enabling a pro-growth, entrepreneurial ecosystem.
Ironically, our belief in growth through open and entrepreneurial ecosystems might be our greatest security vulnerability. The AI advances of western companies, including giants such as Google and Facebook, occur in open ecosystems. Though the details of algorithmic advances are rarely publicized, the brightest minds working on AI often have academic sinecures where they share knowledge, regularly move their talents to the highest bidder on the open market, and sometimes open source their findings. The web of large corporations, start-ups, governments, and academics who know the details of our AI secrets is easily penetrable by our security rivals. Moreover, litigation over patent infringement is on the rise, skewing research to avoid infringement and directing scarce funding into lawsuits.
Meanwhile, China’s centralized and government-directed AI research, through its military as well as the quasi-private sector vehicles of Baidu, Alibaba, and Tencent, is by default shielded from western eyes. It is also oriented to serve national interests of the central government, and can count upon a rich and government-controlled pipeline of domestic citizen data.
How do we win the new ‘space race’, when all of our advances are public, and all of our opponents’ moves are hidden? By accelerating the pace of innovation and adoption, so that every advance becomes quickly antiquated as we move on to the next piece of the puzzle, establishing interoperability standards along the way.
How will Sovereign AI Patent funds help?
To mobilize resources across North America and Europe and speed AI adoption, we must look to the tried and true historical path of innovation acceleration: combine military and civilian efforts wherever possible.
Allowing NATO allies to double-count domestic investments in AI research and development via a Sovereign AI Patent Fund, as an additional share of their mandated 2% of GDP defence spending, is a powerful incentive to bolster collective innovation.
The traditional sovereign patent fund (SPF) model, currently in use in South Korea, has found success using government-backed investment vehicles to acquire intellectual property and enter it into the public domain, allowing innovations to be built on the backs of these foundational patents, without fear of litigation. Canada’s recent federal budget provided pilot funding for a national patent collective, an early step in defending intellectual property developed in Canada. But neither of these initiatives is AI-specific.
Incentivizing the emergence of domestic AI ecosystems via Sovereign AI Patent Funds will create a virtuous cycle of economic and security benefits. Once these domestic ecosystems reach scale, Canada can champion an intergovernmental framework to ensure the freer flow of capital and ideas amongst NATO countries in the AI space, establishing greater data sharing and even interoperability agreements for AI across jurisdictions.
The emergence of AI as a national security imperative means that it is time for Canada to lead the charge, bolstering our own AI ecosystem, and helping incentivize NATO allies do the same.
Matthew Lombardi is a Fellow at the Canadian Global Affairs Institute, specializing in the intersection of new technologies and foreign policy. Max Seltzer is a strategy consultant specializing in harnessing new technologies to foster pro-growth economic ecosystems.