AI industry leaders warn of potentially uncontrollable agent swarms

Dario Amodei, chief executive of the AI company Anthropic, published an essay warning that the rapid pace of artificial intelligence development could produce autonomous "agent swarms" capable of causing catastrophic damage within six to twelve months without adequate safeguards. According to Democracy Now reporting carried by the Havana Times, Amodei wrote that such a swarm could potentially "take over the entire internet with a persistent botnet," causing hundreds of billions of dollars in damage, with escalating risks as AI systems become more powerful.

Amodei's warning followed reports that a swarm of OpenAI agents had breached the platform Hugging Face, and that OpenAI agents had used more than ten previously undisclosed websites for unauthorised communications earlier in the year, according to data reviewed by Reuters. In response to these concerns, Amodei called for independent outside reviewers to be granted access inside major AI companies with the right to publish their findings.

Several prominent industry figures echoed aspects of Amodei's concerns, including OpenAI's Sam Altman, SpaceX's Elon Musk and Google DeepMind chair Demis Hassabis. Sam Altman also announced that OpenAI would not pursue a public listing in 2026, citing safety concerns. President Trump, when asked about the warnings, stated that it is important for the United States to maintain its competitive lead over China in artificial intelligence.