Every Country Is on Its Own on AI: Why AI regulation can’t follow in the footsteps of international nuclear controls
In this Foreign Policy Argument, Bill Drexel, associate fellow at the Center for a New American Security (CNAS) and Michael Depp, a research associate at CNAS argue that, while it is attractive to mirror the precedent of the international nuclear control regime when it comes to AI regulation, similar AI governance structures will not work. First they argue that any institutional approach will likely be unable to keep up with the fast pace of technological developments. Treaties and other multilateral agreements are slow instruments.
Second, compared to the small number of use cases of nuclear technology, AI can be virtually used in all domains of society. Partly because of this enormous potential, many countries see the development of AI increasingly as part of their strategic competitive advantage. However, at the moment states are still figuring out what the risks and opportunities are, and how exactly AI will influence their relative economic and political competitiveness. Until there is more clarity about that impact, it is unlikely that we will see a control regime similar to the one for nuclear technology.