Zuckerberg Rejects Industry-Wide AI Safety Coordination

Mark Zuckerberg said industry-wide coordination on AI safety is unnecessary, arguing each lab should handle risks internally. His stance contrasts with OpenAI, Anthropic and Google DeepMind, which are forming the SAFA standards body and backing the FRONTIER Act.

• Zuckerberg: commercial incentives will push labs to make AI safe
• Meta delayed its Muse release by months over safety concerns
• OpenAI, Anthropic and Google are forming the SAFA standards body by 2027
• OpenAI backs the FRONTIER Act requiring independent AI safety audits