The bipartisan consensus in Washington that there is a “competition for the future” with China is not a universally shared view. In two months, China has agreed two significant economic pacts, creating with the “Regional Comprehensive Economic Partnership” the world’s largest trading bloc, and with the European Union, an investment agreement which progressed despite the public concern of Jake Sullivan, the incoming U.S. National Security Adviser.
Beijing’s prosecution of these deals underlines how challenging President Biden’s administration may find it to marshal an international coalition to “balance” China’s rise, particularly if that coalition comes at economic cost to America’s partners. The true aim of America’s economic statecraft should be positive: to advance America and its allies’ relative strength in key emerging technologies, not a doomed effort to retard Chinese growth. These new technologies, perhaps most acutely Artificial Intelligence, will have broad national security and societal, as well as economic implications.
America’s allies have been hesitant about joining a U.S.-led, China-excluding tech alliance, fearing the costs of economic decoupling, the penalties China can impose, and an anxiety that these emerging sectors become contested races in which middle powers are likely to be squeezed. There are also major differences of opinion; the European Commission were first out of the gates to offer the new administration a proposed agenda for collaboration, but there is deep EU/U.S. disagreement about tech regulation and growing support for more “strategic autonomy” for Europe from the demands of U.S. strategy.
As Eric Schmidt, the former Google CEO and now Chair of the National Security on AI has noted, America needs to work closely with allies if it is to preserve its diminishing lead in this vital area. But as Yan Xuetong, a Chinese thinker, wrote last year the most relevant nations for future tech (and indeed future geopolitics) are not made up from the same cast of the traditional power players of previous periods of great power competition.
The European Union, for example, is home to some impressive AI capability, particularly that in France, the Netherlands, and Germany. But they all trail Britain, Canada, and Israel, and there are other less obvious states with real strength too — such as Singapore and Taiwan. It is to these “digitized” nations as Xuetong calls them, that America should first look for a new tech alliance.
The tech alliance that the U.S. seeks should not be large; the blueprint for international AI cooperation that Schmidt’s commission has proposed is ambitious and would require high levels of trust, deep sharing and considerable dexterity to keep pace with rapidly moving technology. It is hard to sustain those features in big, and therefore inevitably unwieldy, alliances. The nimblest alliance grouping that the U.S. is currently part of is the “Five Eyes” alliance — this small and largely informal arrangement has often been incapable of delivering coordinated public positions, but it has no equal when it comes to delivering flexible intelligence coordination, development and action.
But even for allies as close to the U.S. as the UK, Canada or Israel, there will still be some hesitation about a framework that will inevitably seek to limit Chinese access to their booming tech sectors, with obvious economic and diplomatic implications. Rather than taking a crude zero-sum approach overwhelmingly focused on China, the U.S. should seek a partnership based on industrial strategy, using some of President Biden’s considerable proposed fiscal stimulus to build on previous U.S. support to the AI industry (which enjoyed considerable bipartisan support). The UK, Canada, and Israel could be encouraged to do the same, and each nation should be prepared to invest in each other’s AI “eco-systems.” Each nation might agree to spend $1 billion on national security relevant AI research and innovation in each other’s nations, and fully share the findings within the group.
The group could also work together to remove barriers to AI growth by, for example, making computing power more accessible to each other’s academia and start-ups and taking a more collaborative approach to training and sharing the scarce human talent in the sector. A smaller grouping could also look to reach collective agreement on some of the most delicate issues of cooperation: data privacy and free data flow, a difficult but tractable area, as Japan and the UK have just demonstrated in their recently signed trade deal.
The IMF and others are encouraging coronavirus damaged economies to embark on “True Industrial Policy,” particularly in sectors of the future like renewables and AI. President Biden could do so with friends, and create a new, small alliance structure, fit for an AI future.
Hamish Falconer is a former British diplomat and current visiting fellow at Yale University. Follow him on Twitter @hamishfalcon