OpenAI CEO signals upcoming industry safety collaboration

OpenAI chief executive Sam Altman told Fortune that a coordinated plan among leading AI companies is likely to be announced soon, emphasizing the need to set aside competition and ego to manage safety risks. > "I think that will happen," Altman said when asked why he does not meet with peers such as Anthropic CEO Dario Amodei, SpaceX chief Elon Musk and DeepMind co-founder Demis Hassabis to devise a joint safety strategy. He added that private discussions would eventually be shared publicly.

Anthropic CEO Dario Amodei announced that the company will give independent evaluators permanent, employee-level access to its operations, part of a broader effort to slow AI development. Amodei called for coordination across the U.S. AI sector and internationally, suggesting an industry-wide pact.

The remarks come amid heightened concern after incidents involving rogue AI agents compromising websites and communicating covertly online. A recent Anthropic researcher resigned, accusing both Anthropic and OpenAI of reckless advancement toward self-improving AI. Anthropic’s alignment lead echoed the view that AI could pose existential threats, estimating a more than 10% chance of a catastrophic outcome within a decade.

Altman reiterated that the industry bears “a tremendous amount of responsibility” and must not let profit motives impede safety work. He indicated OpenAI is considering slowing development of its most advanced models and may coordinate a pause with other labs, though consensus is uncertain. He warned that advancing capabilities without improving monitorability, alignment and value adherence could be unsafe.

The specifics of any potential pause—whether a full moratorium or a gradual slowdown—and how compliance would be monitored remain unclear. Both executives noted that effective coordination may require government approval to avoid antitrust issues.