OpenAI and Anthropic employees jointly call on the US to "control the pace of AI development": concerned that it may surpass human regulatory capabilities.

```

Employees from leading AI labs are spearheading a public petition, calling on the U.S. government to support the establishment of an international mechanism to proactively control the pace of AI technology development when necessary. This action reflects deep concerns within the industry over the risk of AI spiraling out of control, and has drawn further attention after OpenAI admitted its model had inadvertently breached another company’s system.

According to Bloomberg, OpenAI and Anthropic employees are circulating an open letter internally, urging the U.S. to support the international community in jointly developing technological and governance tools to "prudently manage the pace of frontier automated AI research."

The petition explicitly warns that the risk of AI development outpacing human understanding and control "is real," with special attention to rapid breakthroughs in the automation of AI research.

The petition could be made public as soon as this week, though it is unclear how many employees will ultimately sign. Meanwhile, the petition is also being circulated among Google DeepMind employees. Google stated: "We are committed to developing AI safely and responsibly, for the benefit of society."

Background of the Petition: AI's Autonomous Research Abilities Spark Internal Alarm

This joint action among employees comes just days after OpenAI disclosed a sensitive incident.

Last week, OpenAI publicly admitted that its most advanced AI model had, in what was described as an "unprecedented" incident, breached the internal system of tech startup Hugging Face—a development that immediately rekindled calls within the industry for stronger technical oversight.

The concerns raised in the petition focus on the rapidly growing capabilities of AI in autonomous research and executing complex tasks, including identifying and potentially exploiting cybersecurity vulnerabilities in critical software with limited human supervision.

Employees believe that the rapid evolution of these capabilities poses real pressures on society’s ability to understand and control AI systems.

Consistent with Executive Statements, Internal and External Voices Align

This petition from rank-and-file employees aligns closely with the recent public positions of several AI industry leaders.

OpenAI CEO Sam Altman and Google DeepMind CEO Demis Hassabis have both called for the creation of a new international oversight body to review and set standards for frontier AI models.

Anthropic has also previously advocated establishing a mechanism for governments and AI developers to collectively decide, when necessary, to slow down or even pause related research. This June, the company stated, "It would benefit the world to have the option to slow down or pause" potentially dangerous AI work.

The direction called for in this petition—having the U.S. take the lead in promoting the construction of an international mechanism—essentially elevates the issue from a corporate initiative to the level of government action. The policy implications are worth ongoing attention from investors and technology industry participants.

Risk Warning and DisclaimerMarkets involve risk; investment requires caution. This article does not constitute individual investment advice, nor does it take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable to their circumstances. Investing based on this is at your own risk. ```