Zuckerberg says Meta delayed shipping Muse for safety

The debate over AI safety has intensified as prominent labs call for slowing AI advancement. Dario Amodei's nearly 4,000-word essay argued for decelerating development to deploy adequate guardrails, involving international collaboration between companies and governments. This call was endorsed by rivals like OpenAI's Sam Altman and xAI's Elon Musk, but not everyone agrees.

Meta CEO Mark Zuckerberg took to X this week to share his view that trust and alignment are becoming key differentiators for AI agents and models. He revealed that Meta delayed shipping its AI model Muse for several months to focus on safety and security, framing it as part of normal work rather than a response to external calls. While endorsing parts of Amodei's plan, Zuckerberg seemed to imply government action is unnecessary and that market incentives will compel companies to prioritize safety.

Other executives weighed in. Reddit co-founder Alexis Ohanian told CNBC the industry had been 'tone deaf' on explaining AI risks but hoped companies would figure it out themselves. Google DeepMind co-founder Shane Legg said capabilities are advancing quickly but 'we can't let capabilities get ahead of safety.' Meanwhile, a report from The Information said OpenAI, Anthropic, and other major AI companies were working together to create an AI standards organization — described as a private 'self-regulatory body.' This comes as the Trump White House and many in Congress show little interest in regulating AI, with 'AI czar' David Sacks saying regulation should be left to companies. Critics, including China's Ministry of Foreign Affairs and Cohere CEO Aidan Gomez, accuse big American AI labs of seeking regulatory capture to disadvantage smaller competitors. Gomez called it a 'cartel' dispute over who writes the rules. The outcome hinges on whether voluntary industry standards are seen as credible or self-serving, and how they affect competition.