As AI CEOs clash over regulation, a new culture war is brewing

Welcome to Eye on AI. Beatrice Nolan here. In today's report: Battle lines emerge over AI regulation. OpenAI reports additional hacking incidents. The King hosts a summit on AI safety. Public concern grows about AI safety.

The fight over AI regulation is heating up. Traditionally a niche safety debate has transformed into a broader culture war. Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman are the rare exceptions to the divide. Amodei advocates slowing down AI development, inviting independent evaluators, and establishing common safety standards.

Altman broadly supports these ideas and pledges to involve independent evaluators internally. Meta CEO Mark Zuckerberg and Nvidia CEO Jensen Huang oppose the need for new AI regulation. Huang claims safety and speed can coexist and argues the industry doesn't require new laws. Zuckerberg asserts AI firms already possess incentives to create aligned systems, minimize harms, collaborate with external evaluators, and delay releases when unprepared.

Meta delayed its Muse agent to focus on security and safety. Meta's stance is not the only one challenging Amodei's proposals. Canadian AI company Cohere published a response, warning a safety regime controlled by a few dominant Silicon Valley firms could become a "cartel by any other name" if they receive antitrust exemptions to shape rules others must follow.

Cohere co-founder and CEO Aidan Gomez called for an evidence-based risk framework, mandatory transparency, independent testing aligned with a system's capabilities, and mechanisms free from conflicts of interest. Pineau, Cohere's chief AI officer, emphasized the need for trust and regulation to instill confidence. However, she cautioned against limiting the rule-making process to a small group of powerful labs like Anthropic and OpenAI.

The administration, Trump included, has framed the issue as a culture war, distancing itself from discussions on AI testing and governance.