An AI Oversight

Ian Bremmer and Mustafa Suleyman’s essay (“The AI Power Paradox,” September/October 2023) warns of the dystopian future that will arrive thanks to artificial intelligence. The authors write that policymakers around the world have only “begun to wake up to the challenges posed by AI and wrestle with how to govern it” with the “Hiroshima AI process,” a G-7 initiative launched in May 2023. But Bremmer and Suleyman ignore a huge amount of relevant global policy work on AI.

The Hiroshima process is only the most recent AI policy initiative of the past decade, with Japan having led a global effort to regulate AI since the 2016 G-7 summit. Other initiatives include the widely endorsed 2018 Universal Guidelines for AI, supported by scientific societies and human rights experts; the Organization for Economic Co-operation and Development’s 2019 AI Principles; the G-20’s 2019 AI Guidelines; UNESCO’s 2021 Recommendation on the Ethics of AI; the EU’s forthcoming AI Act; and the Council of Europe’s forthcoming AI Treaty. There is no discussion of any of these landmark events in AI governance in Bremmer and Suleyman’s essay. As a result, instead of having us build on prior efforts, the authors would send us hurtling back to 2015, before governments had begun working together on global AI policy. This approach by Bremmer and Suleyman makes the problems of AI governance appear more intractable than they are. AI policymaking is an evolutionary process that requires paying attention to how governments are already meeting emerging challenges in technology.

Marc Rotenberg, Founder and Executive Director, Center for AI and Digital Policy