Anthropic CEO Calls for Slower AI Development Amid Rising Safety Concerns

Anthropic CEO Dario Amodei is calling on AI companies to slow the pace at which they develop increasingly capable artificial intelligence models.

According to Reuters, Amodei made the case in an essay published on September 12, arguing that AI progress is moving quickly enough that companies need to create more time for safety work, independent evaluation and coordination.

The comments come as the AI industry continues to develop increasingly capable models and autonomous AI agents.

Why Anthropic Wants a Slower Pace

Amodei argued that the industry should not stop AI development, but should deliberately create more time to understand and manage the risks associated with increasingly capable systems.

One concern he highlighted is the possibility of AI systems becoming better at improving future AI systems. He argued that if this process develops faster than researchers can understand and control it, existing safety practices may not be sufficient.

The discussion is part of a broader debate over whether AI capabilities are advancing faster than safety testing and governance.

A Three-Step Approach to AI Safety

Amodei proposed a three-step framework for managing the pace of advanced AI development.

The approach includes stronger independent evaluation, greater coordination between AI companies and governments, and broader international cooperation.

One proposal would give independent evaluators ongoing access to AI companies' internal systems and safety assessments so they can better evaluate how advanced models behave.

Why This Matters Beyond Anthropic

The debate is not limited to Anthropic or Claude.

AI companies across the industry are developing models that can write software, analyze large amounts of information, use external tools and perform increasingly complex sequences of tasks.

That means questions about testing, access controls, monitoring and human oversight are becoming increasingly important for the entire AI ecosystem.

The issue is ultimately about finding a balance between technological progress and the safeguards needed to deploy increasingly capable systems responsibly.

Why This Matters for the Future of AI

The debate over the pace of AI development is becoming increasingly important as companies compete to build more capable models and autonomous systems.

Whether the industry ultimately adopts slower development schedules remains uncertain. However, calls for stronger safety evaluation show that AI governance is becoming a central part of the technology conversation.