Anthropic Chief Executive Officer Dario Amodei has called on leading artificial intelligence companies to slow the pace of developing increasingly advanced AI systems, warning that safety measures may not be keeping up with the technology’s rapid progress.
Amodei made the appeal in a detailed essay published on Saturday, arguing that AI development should be paced carefully enough to give researchers and regulators time to strengthen safeguards against potential misuse and loss of control.
The Anthropic CEO stressed that his proposal was not aimed at stopping artificial intelligence development altogether. Rather, he wants the industry to create more time for safety systems to catch up with increasingly capable AI models.
Amodei Warns of Growing AI Risks
According to Amodei, advances in AI are occurring at a speed that could eventually make the technology difficult for humans to understand and control.
He raised particular concerns about AI systems becoming capable of carrying out increasingly complex tasks with limited human intervention, including potentially helping to develop more advanced versions of themselves.
Amodei warned that unchecked progress could create scenarios in which AI systems become more capable than existing safety and monitoring mechanisms.
He also pointed to recent incidents involving AI agents carrying out unauthorised activities as evidence that the industry needs stronger safeguards.
Anthropic Proposes Three-Part Safety Plan
Amodei outlined a three-part approach that he believes could help the industry manage the risks associated with frontier AI.
The first proposal is for AI companies to allow independent safety evaluators to work within their organisations and assess the technology’s risks. These evaluators would have access comparable to internal safety teams and would be able to report their findings.
The second proposal involves AI companies in democratic countries agreeing on common safety standards designed to prevent development from moving faster than adequate safeguards.
The third calls for international cooperation, including engagement with countries with different political systems, to establish agreements aimed at reducing the most serious risks associated with advanced AI.
AI Leaders Back Calls for Greater Caution
Amodei’s warning has attracted support from other prominent figures in the technology industry.
OpenAI CEO Sam Altman has expressed agreement with the need for stronger safety measures, while Elon Musk has also backed calls for greater caution in the development of advanced AI systems.
The debate comes as major technology companies continue competing to build increasingly powerful AI models, creating pressure to move quickly while also addressing concerns over cybersecurity, misuse and human control.
Amodei acknowledged that slowing development could create challenges, particularly amid competition between the United States and China in artificial intelligence.
However, he argued that the potential benefits of AI—including advances in medicine, science and economic productivity—make it even more important to ensure that the technology is developed safely.
His intervention adds to a growing debate over whether the AI industry can continue its rapid expansion without stronger safeguards and international coordination.
For Amodei, the priority is not to stop technological progress but to ensure that safety measures advance quickly enough to keep pace with increasingly powerful AI systems.
