Anthropic Calls For International Coordination On Advanced AI

What happens when a leading player asks to slow down a race they themselves help to accelerate? Anthropic, the company behind the chatbot Claude, believes artificial intelligence could soon cross an unprecedented threshold: that of a technology capable of helping design its own successors. Facing this prospect, the company calls for a coordinated pause in the development of the most advanced systems, reigniting the debate on the ability of states and companies to retain control over AI.

Anthropic worries about the emergence of an AI capable of improving its successors

This reflection is based on developments already observed internally. Anthropic claims that Claude now plays a major role in its own development, as the firm has just opened its chatbot to 150 organizations. According to the company, about 80 % of the new code used to improve the model is now generated by Claude itself.

Researchers clarify that current systems are not yet capable of independently developing a new generation of AI. However, they highlight several advances that fuel their concerns :

• Claude already generates a large portion of the code used to improve its future versions ;
• AI agents are now able to write and execute code autonomously ;
• Some systems can delegate tasks to other specialized agents ;
• The speed of progress could become largely dependent on the available computing power if recursive self-improvement became a reality.

These capabilities do not yet prove that an artificial intelligence could design its successor alone. However, they illustrate the technological trajectory prompting Anthropic to anticipate this scenario and to now reflect on the control mechanisms that might become necessary.

The call for global coordination reignites the debate on AI governance

Beyond the technical dimension, Anthropic questions the mechanisms that would allow reaction if such a technological breakthrough occurred. Using on-chain data, the company speaks of the potential need for a temporary and coordinated pause in the development of the most advanced systems to give authorities and researchers time to assess the risks. It believes such a measure could only work within an international framework. Without coordination, actors continuing their work could quickly make up for lost ground and neutralize the effect of a voluntary slowdown observed elsewhere.

This position already triggers mixed reactions. It fuels a debate between those who see this warning as a prudent anticipation of a real risk and those who view it as a form of communication aimed at raising public awareness about the stakes of advanced artificial intelligence. The issue now goes beyond the researchers’ circle and also concerns governments, regulators, and large companies engaged in this technological competition.

Discussions about artificial intelligence no longer focus solely on its performance, but also on measures capable of framing its evolution. By calling to reflect today on a global slowdown mechanism, Anthropic reveals an unprecedented challenge. The question remains whether states and companies will be able to agree on common rules before AI progress makes this debate even more urgent. For now, the European Union maintains its regulatory schedule without yielding to pressures.