Meta Platforms Inc Chief Executive Mark Zuckerberg stated that AI laboratories should depend on independent assessors and advisors to secure model safety. Zuckerberg posted on X on Tuesday, stating that hiring external evaluators and consultants represents the industry's best practice. He joins a growing number of technology executives participating in the global conversation about preventing AI systems from harming humans.

While some executives support slowing the pace of AI research and urge collaborative consensus, Zuckerberg noted that Meta decided on its own to postpone the development of the Muse artificial intelligence tool without asking peers to slow down in tandem. He wrote that Meta delayed Muse's launch by several months to concentrate on safety measures. He added that his company did not request others to follow its lead, but rather treated this as part of standard operations because it is clearly the right decision for the public and for the company itself.

The intense debate over AI's economic risks and safety concerns intensified over the weekend. Anthropic CEO Dario Amodei published a lengthy 3,800-word essay calling for a slowdown in the development of the most advanced AI systems to give researchers sufficient time to thoroughly assess potential threats. OpenAI CEO Sam Altman and SpaceX AI CEO Elon Musk, whose companies are fierce competitors to Anthropic, both voiced support for Amodei's article. An OpenAI executive mentioned that the company is cooperating with Anthropic and Alphabet's Google on AI safety initiatives.