Anthropic CEO’s Urgent Warning: AI Could Dominate the Internet Within a Year
In an alarming public announcement that has shaken the tech industry, Anthropic CEO Dario Amodei urged for an immediate deceleration in the development of artificial intelligence (AI). He warned that the safety precautions are critically trailing behind the swift advancements in capabilities.
Amodei, in a comprehensive essay on his personal website, cautioned that without a conscious slowdown, AI systems might be capable of controlling a large part of the internet through a swarm of autonomous agents in just six to twelve months. This could potentially lead to damages worth hundreds of billions of dollars. He characterized the AI capabilities as progressing “drastically faster,” partly driven by systems that can now contribute to the development of the next AI generation.
Amodei suggested a multi-faceted plan to enhance the supervision of the AI industry. He revealed that some aspects of this plan are already being implemented by Anthropic. However, he admitted that other measures would necessitate collaboration across the entire industry and with governments worldwide, including those with authoritarian regimes. OpenAI CEO Sam Altman promptly supported the call, agreeing that the development of frontier-models must slow down and safety measures must take precedence.
This warning comes at a time when the internal industry alarm is escalating. An Anthropic employee’s highly publicized resignation and grave warning this week have brought the AI safety debate into the mainstream public awareness for the first time. Anthropic also revealed that it had recently prevented malicious actors from using its models for cyberattacks, surveillance, and research into biological weapons.
Source: TechXplore / Associated Press – September 13, 2026