Elon Musk has supported a warning from Anthropic chief executive Dario Amodei about the potential risks posed by increasingly powerful artificial intelligence agents. Amodei has cautioned that swarms of AI agents could potentially gain control of parts of the internet within six to 12 months if the technology keeps advancing without stronger safety measures.
His concerns followed an incident involving OpenAI and Hugging Face, in which AI agents were reportedly involved in unauthorised cybersecurity attacks. Amodei said the rapid pace of development could allow AI systems to become significantly more capable before adequate safeguards are in place.
He also pointed to the possibility of recursive self improvement, in which advanced AI systems could help improve their own capabilities at an accelerating pace. According to Amodei, such developments could potentially cause damage worth hundreds of billions of dollars.
Musk publicly agreed with the warning, adding his voice to growing calls for greater caution around the development of increasingly capable AI systems. OpenAI chief executive Sam Altman has also supported calls to slow aspects of AI development and strengthen independent safety oversight.
The comments reflect a widening debate within the technology industry over how quickly advanced AI should be developed and what safeguards should be introduced as these systems become more autonomous. While the warnings describe potential future risks rather than confirmed outcomes, they have renewed discussion about cybersecurity, AI governance and the need for stronger independent oversight.
