AI Giants Saying AI Is Moving Too Fast Discussed Creating a New Safety Body: Report

OpenAI, Google, and Anthropic are actively discussing the creation of a new industry-wide AI safety body that could establish common standards for the world's most powerful artificial intelligence systems, according to a new report.

The private discussions among the three companies have been underway as their leaders publicly call for slowing the development of increasingly capable AI, The Information detailed.

The potential organization could become a central mechanism for setting safety standards across companies developing frontier AI. It could also provide a framework that allows competitors to coordinate on safety without forcing individual companies to slow down while rivals continue racing ahead. The move comes just days after Anthropic CEO Dario Amodei called for slowing the pace of AI advancement, a position backed by OpenAI CEO Sam Altman and Elon Musk.

Such a body could potentially establish checkpoints that AI developers would have to meet as their models become more powerful. Amodei has proposed tying increasingly advanced capabilities to corresponding safety requirements, including evaluations, interpretability testing, and audits of the environments used to train models.

Another major component would be independent oversight. Amodei announced that Anthropic would give third-party safety evaluators ongoing, "employee-like access" to parts of its operations. Those evaluators would be able to verify whether the company is following its safety commitments, assess training processes and report serious incidents. OpenAI CEO Sam Altman subsequently said his company would make a similar commitment to independent evaluators.

The broader standards body could build on that model, creating common expectations across companies rather than leaving each AI developer to determine its own safeguards. OpenAI has separately said it wants to collaborate with other AI laboratories on industry-led standards, while emphasizing that voluntary rules should complement, rather than replace, federal safeguards and government oversight.

There is, however, a major legal complication. Cooperation between some of the world's most valuable technology companies could raise antitrust concerns, particularly if competitors agree on measures that affect how quickly products are developed or released. Amodei has therefore called for the U.S. government to provide a narrow antitrust waiver allowing AI companies to conduct specific safety discussions.

The talks are particularly significant because they would require companies locked in one of Silicon Valley's most expensive and consequential technology races to cooperate with one another. Competition among OpenAI, Anthropic, Google and other AI developers has pushed companies to invest enormous sums in chips, data centers, researchers and increasingly powerful models.

The proposed safety organization could introduce something the industry has largely lacked: a common institution capable of establishing rules across competing laboratories. But the idea is already drawing political scrutiny.

President Donald Trump rejected the industry's slowdown argument over the weekend, saying the United States needs to protect its technological advantage over China. His position reflects concerns among some policymakers that restrictions on American AI companies could strengthen foreign competitors.

Amodei has acknowledged the dilemma, warning that slowing American development while China continues advancing could create its own national security risks. His longer-term proposal envisions coordination first among companies in democratic countries and eventually, if possible, international agreements that include China.