Rival AI Companies Discuss Shared Safety Standards
OpenAI is holding discussions with major artificial intelligence competitors, including Anthropic and Google, about possible cooperation on industry safety standards, according to a company spokesperson. The talks follow a July proposal from Google DeepMind chair Demis Hassabis, who called for a United States-led standards organization combining private-sector participation with federal oversight.
Hassabis suggested a structure resembling the Financial Industry Regulatory Authority, a self-regulatory organization that operates under government supervision. The proposal comes as leading AI developers face growing scrutiny over the safety and security implications of increasingly capable models.
Although OpenAI, Anthropic and Google compete directly in advanced AI development, their leaders have recently expressed support for coordinating on risk management. Anthropic CEO Dario Amodei has urged companies to reduce the pace at which they improve their most advanced systems. Hassabis described that position as broadly appropriate, while noting that implementation details remain unresolved. OpenAI CEO Sam Altman also said the issue has become a central topic within his company.
OpenAI has separately outlined plans to promote voluntary industry-led standards in parallel with government action. Chris Lehane, the company’s chief global affairs officer, said such measures should supplement rather than replace federal safeguards and democratic oversight. He added that companies may begin building a cooperative framework regardless of whether government support is immediately available.
The discussions indicate that leading AI firms are considering both public regulation and private coordination as potential responses to emerging risks. However, no formal agreement, governing structure or implementation timetable has been announced. Google and Anthropic had not commented on the reported talks when CNBC published its report.
The effort remains preliminary, with questions about participation, enforcement, federal oversight and whether voluntary commitments can keep pace with technological development. The companies have not announced a timetable for further action.