Blog

5 AIMS Become Visible at Paris AI Summit

Posted on 12/02/25 in AI, Politics

In recent weeks, Wendy Hall and I have begun to adapt our ideological/geopolitical view of Internet governance, which we expressed in our book Four Internets, to the governance of AI, publishing short pieces with Sciences Po and the Atlantic Council. Broadly speaking, we have seen an analogous pattern emerge, with existing attitudes carrying over into five Artificial Intelligence Management Strategies, which we call 5 AIMS.

The 5 AIMS each encapsulate a basic view of the good or bad effects of technology. They can be mixed and matched, combined, combated or denied, but they can’t be eliminated. The Open AIMS valorises openness and innovation; the Bourgeois AIMS rights; the Paternal AIMS safety and protection; the Commercial AIMS market solutions to collective action problems; and the Hacker AIMS the power of the individual coder to undermine centralisation and self-declared authority.

After a series of AI get-togethers which usually produced outcomes of harmony and agreement (if not necessarily clarity), the French/Indian AI Summit in Paris ended more rancorously, with the US and the UK refusing to sign the communiqué and its exhortations to ensure that “AI is open, inclusive, transparent, ethical, safe, secure and trustworthy, taking into account international frameworks for all” and to make it “sustainable for people and the planet”.

Maybe there is a limit to the amount of motherhood and apple pie an industry can consume. But what we are really seeing is the friction between the 5 AIMS which previous summiteers have managed to suppress. The EU approach in legislation and rhetoric has focused on the Bourgeois AIMS, on the rights of citizens, with the addition of some concern about safety, which brings in the Paternal AIMS with its risk-based approach. Paternalism is an approach on which many can agree, because safety concerns, even when genuine, are a useful cover for control. China has long been a paternalist regulator of digital technologies, and its mantra of preserving social stability is popular in the country as a whole, with the agreeable side-effect of delegitimising attacks on the Communist Party. Hence China’s signing of the communiqué, to the surprise of some, but consistently with the 5 AIMS view.

However, there have long been worries that a focus on rights and safety will undermine innovation and raise compliance costs for start-ups, concerns of both the Open AIMS and the Commercial AIMS. It is perhaps not a surprise that it took a senior representative of the Trump administration to shatter the illusion of consensus. The administration itself may be deeply conflicted between its MAGA devotees, who are chiefly interested in undermining the state and promoting the independence of citizens, and its tech bros, who basically want to make money, but all in this perhaps shaky coalition agree that “ethical, safe and sustainable” stand in the way of “effective, productive and valuable”. The British reasons for declining to sign the communiqué were less clear, but a one reason may have been to disavow Britain’s historical role in outlining the safety agenda.

But another agenda is also becoming clearer: AI nationalism. The promotion of domestic AI industry is of a piece with a view of AI as of strategic importance. The collapse of the rule-based order, and the rise of the civilisational state have resulted in the spread of the view that this is a zero sum game. That really a variant of the Paternal AIMS safety view (safety of the nation in a hostile world, rather than of the individual citizen), but it is now a widespread concern, including in America.

This was clear in J.D. Vance’s anti-China remarks in Paris – interpreting the Paternal AIMS not as a means to protect citizens against hypothetical dangers, but rather to advance the interests of the nation. His warning against authoritarians was pointedly not a criticism of authoritarianism, merely a suspicion of cooperation and openness. “Some of us in this room have learned from experience partnering with [authoritarian states] means chaining your nation to an authoritarian master that seeks to infiltrate, dig in and seize your information infrastructure.” India’s adoption of the Hindutva cultural ideology in the Modi years has similarly had the effect of promoting paternalism, and we have already noted China’s view. In transactional geopolitics, openness goes as far as the border, no further.