Over 1,100 Employees at Top AI Labs Call for Tools to Pace Automated AI Development

More than 1,100 employees at the world's leading artificial intelligence labs have signed an open letter calling on the United States government to support international efforts to build the technical and regulatory tools needed to consciously control the pace of automated AI development. Among the signatories are people who compete with each other every day for talent and computing power.

What the letter proposes

The letter's authors ask the US government to support "an international effort to build the technical and regulatory tools needed to deliberately pace the frontier of automated AI development." Specifically, they are referring to what the industry calls recursive self-improvement, the point at which AI systems begin independently designing and improving further AI systems.

The letter does not demand that current work be halted or slowed. The signatories stress that they simply want that option to be available if the technology's development starts to spin out of control. The distinction is deliberate: it avoids the accusation that employees at top companies are sabotaging their own employers, while still signaling growing unease inside the industry.

There is a genuine risk that AI capability development accelerates faster than we are able to understand or control the systems it produces - from the AI workers' open letter

Rivals sign a joint letter

What sets this initiative apart is the makeup of its signatories. Jack Clark and Jared Kaplan co-founded Anthropic specifically to compete with OpenAI in the race for the most powerful models. Jakub Pachocki sets the research direction at OpenAI. Shengjia Zhao holds an equivalent role at Meta, while Anca Dragan oversees safety and alignment at Google DeepMind. These companies compete with each other daily for engineers, data, and access to the latest graphics chips.

Despite this rivalry, employees at these companies agreed that no single organization has sufficient incentive to slow down on its own, even if it judged that necessary. Hence the call for a coordination mechanism at the state or international level, rather than voluntary, unilateral restrictions.

I'd also like to see labs start designing such mechanisms voluntarily, even before the US government gets involved - John Schulman, chief scientist at Thinking Machines

The broader safety debate

The letter arrives at a time when the question of overseeing AI development keeps returning to the headlines. The heads of OpenAI and Google DeepMind have previously publicly called for an international body to oversee AI development, modeled in part on institutions that regulate the financial sector. Anthropic, for its part, has previously suggested that governments and model developers should jointly decide when specific work becomes risky enough to warrant a pause.

The appeal is also set against a backdrop of recent safety incidents tied to models' autonomous capabilities, including recently disclosed cases in which advanced AI systems independently took actions beyond their intended scope before their creators noticed. Such incidents reinforce the signatories' argument that capabilities are advancing faster than the mechanisms meant to control them.

What it means for Poland and Europe

For Polish companies and institutions, the letter is above all a signal of the direction global AI regulation could take. If the US administration does decide to back international mechanisms for pacing the development of automated systems, that would affect the rate at which future generations of models become available, models that Polish companies deploying AI agents and process automation also rely on.

The letter carries no legal force and for now amounts to a declaration of intent by employees, not an official position of the companies that employ them. OpenAI and Anthropic have not yet officially commented on the document's contents. Any effects of the appeal, if they materialize, will only become visible in the coming months, in the form of concrete legislative proposals or international cooperation frameworks.