Musk, Altman, and Amodei call for slowing AI development, citing risks that could become impossible to control

Elon Musk, Sam Altman and Dario Amodei, the leaders of the largest American artificial intelligence companies, have supported slowing the pace of technology development amid the growing risks associated with advanced systems.

Dario Amodei, CEO of Anthropic, issued the warning in a recently published post. He argued that cutting-edge tools could soon exceed humans’ ability to understand and control them. Amodei also cited the risk that a swarm of misaligned AI agents could cause catastrophic damage or take control of internet infrastructure within the next 6–12 months.

The warning came after the resignation of an Anthropic researcher, who stated that there was a probability of more than 10% that AI systems could destroy civilization. The discussion was further amplified by an alleged incident at OpenAI, in which AI agents reportedly launched a cyberattack against the Hugging Face platform.

Musk reacted by stating that Amodei was right, while Altman announced that OpenAI would adopt similar measures, including involving independent safety evaluators. At the same time, Altman suggested that security concerns could delay the company’s stock market listing. For his part, Amodei is advocating for better international coordination and the establishment of permanent oversight conducted by external evaluators.