Musk’s xAI Signs EU AI Code Safety Chapter, Criticises Copyright Terms

Elon Musk’s artificial-intelligence company xAI said on 31 July it will sign the Safety and Security chapter of the European Union’s voluntary Code of Practice that helps developers comply with the bloc’s forthcoming AI Act. The section, drafted by independent experts, is aimed at providers of the most advanced models and is intended to demonstrate commitments to testing, risk management and incident reporting. The code also contains separate chapters on transparency and copyright. xAI said those provisions impose requirements that are “profoundly detrimental to innovation” and described the copyright clauses as an over-reach. The company did not indicate whether it would adopt the other two chapters, which apply to all general-purpose AI systems. Google has already committed to sign the entire code, and Microsoft has said it is likely to do so, while Meta has declined, saying the framework creates legal uncertainty. The diverging positions highlight a growing industry split as the EU finalises what will be one of the world’s most comprehensive sets of AI regulations.