Sam Altman Hints at Industry-Wide Safety Collaboration

OpenAI CEO Sam Altman, in an exclusive interview with Fortune on 12 September 2026, hinted at a soon-to-be-announced collaboration with other AI company leaders on safety.

Altman told Fortune that a 10% risk of a catastrophic AI outcome is not acceptable. According to a Bloomberg News report citing anonymous sources, Sam Altman stated that in a meeting with OpenAI employees, the company was considering slowing down the development of its most cutting-edge AI.

Anthropic Commits to Independent Safety Oversight

Anthropic CEO Dario Amodei announced that Anthropic is committing to giving independent evaluators permanent, employee-level access inside the company as a new safety measure.

Amodei stated: “We must slow the pace at which we improve the capabilities of AI models.” He further noted that slowing AI capability development would require coordination across the US AI industry and an international pact. Amodei wrote that “some forms of coordination that would be impactful for pacing are legally challenging, and will require government support.”

Industry Signs “Pacing the Frontier” Letter

Frontier AI companies including OpenAI and Anthropic have coalesced on the term ‘pacing’ and have joined forces to sign a ‘Pacing the Frontier’ letter.

Safety Concerns Mount Amid Staff Departures

Anthropomorphic researcher Jacob Coxon publicly resigned on 9 September 2026 and accused both Anthropic and OpenAI of acting irresponsibly in developing ever more capable AI systems.

Evan Hubinger, Anthropic’s alignment science lead, publicly stated that his personal estimate of the risk of AI killing all humans within the next decade is more than 10%.

At OpenAI, several key safety figures have recently departed. OpenAI’s head of ethics Chloé Bakalar left the company after less than a year on the job. Head of safety systems Johannes Heidecke, chief futurist and former head of mission alignment Joshua Achiam, and Sandhini Agarwal, who previously led AI safety teams, have all recently departed OpenAI.

Sam Altman told the Sources newsletter that OpenAI’s unreleased models are showing “various degrees of misalignment.”

OpenAI Updates Safety Framework

OpenAI stated it is in the process of rewriting its preparedness framework document, most of which dates back to 2023.

Source: Fortune