More than 1,100 employees from Meta Platforms, Anthropic, OpenAI and Alphabet's Google urged the U.S. government on Tuesday to back an international effort to create technical and governance tools for controlling the pace of advanced AI development. The initiative, called Pacing the Frontier, asks Washington to support work that can “deliberately pace the frontier of automated AI development,” as AI companies build increasingly powerful systems and concerns grow that safety, governance and oversight mechanisms may lag behind.

Pacing the Frontier is supported by nonprofits Guidelight AI Standards and Encode AI. Signatories include Anthropic CEO Dario Amodei and several cofounders, Meta vice president of AI research Dawn Song, and OpenAI Chief Scientist Jakub Pachocki. The statement says the needed tools should help manage how quickly frontier AI systems advance, rather than leaving the speed of development only to company competition.

OpenAI responded on X that “at some point in the future,” acceleration in frontier model development may become high enough that the world will need to pace AI advancement. Anthropic said on X that its research on recursive self-improvement (AI improving its own abilities), published last month, points to the need for tools that let society prepare. Last month, the Claude Code creator called on major AI labs to consider a coordinated pause, warning that AI systems could soon improve themselves faster than society can manage the risks.

Separately, Nvidia said on Monday that it had formed a coalition with Adobe, CrowdStrike and others to build AI safety and cybersecurity tools. The move followed a recent Hugging Face cybersecurity incident that raised concerns over autonomous AI agents (software acting toward goals with limited supervision).

More than 1,100 employees from Meta Platforms, Anthropic, OpenAI and Alphabet's Google urged the U.S. government on Tuesday to back an international effort to create technical and governance tools for controlling the pace of advanced AI development. The initiative, called Pacing the Frontier, asks Washington to support work that can “deliberately pace the frontier of automated AI development,” as AI companies build increasingly powerful systems and concerns grow that safety, governance and oversight mechanisms may lag behind.

Pacing the Frontier is supported by nonprofits Guidelight AI Standards and Encode AI. Signatories include Anthropic CEO Dario Amodei and several cofounders, Meta vice president of AI research Dawn Song, and OpenAI Chief Scientist Jakub Pachocki. The statement says the needed tools should help manage how quickly frontier AI systems advance, rather than leaving the speed of development only to company competition.

OpenAI responded on X that “at some point in the future,” acceleration in frontier model development may become high enough that the world will need to pace AI advancement. Anthropic said on X that its research on recursive self-improvement (AI improving its own abilities), published last month, points to the need for tools that let society prepare. Last month, the Claude Code creator called on major AI labs to consider a coordinated pause, warning that AI systems could soon improve themselves faster than society can manage the risks.

Separately, Nvidia said on Monday that it had formed a coalition with Adobe, CrowdStrike and others to build AI safety and cybersecurity tools. The move followed a recent Hugging Face cybersecurity incident that raised concerns over autonomous AI agents (software acting toward goals with limited supervision).