OpenAI backs binding UK AI legislation

Key Points:

• OpenAI’s head of policy in Europe, Tom Duff Gordon, urges the U.K. to implement binding, capability-based AI safety regulations to address the risks posed by rapidly advancing AI technologies.
• Duff Gordon emphasizes that legislation should focus narrowly on frontier AI models and labs, incorporating independent third-party testing, cybersecurity measures, and mandatory safety incident reporting.
• The call for regulation comes amid growing political pressure on the Labour government, which has so far favored sectoral regulation and voluntary safety partnerships over comprehensive AI laws.
• OpenAI supports the U.K. adopting rules that align with emerging EU and U.S. frameworks to ensure international interoperability and maintain its leadership in AI safety governance.
• Duff Gordon highlights the importance of global dialogue on AI safety, noting upcoming discussions between U.S. and Chinese leaders, and sees the U.K. as positioned to play a key role in shaping future AI policy.