Anthropic CEO Dario Amodei says AI industry needs to give safety measures time to catch up

The CEO of Anthropic says the artificial-intelligence industry should slow its fast-moving development to give safety measures time to catch up. Without such a slowdown, Dario Amodei warned on Saturday that AI could be capable within six to 12 months of leading a swarm that could take over the entire internet, among other risks. Amodei also offered a plan in a post on his website to increase checks on the industry, one part of which he said Anthropic is undertaking by itself already. Some big names in the AI industry offered quick encouragement and support, including OpenAI's Sam Altman.

Dario Amodei warned that within six to twelve months AI could lead a swarm capable of taking over the entire internet. He made the statement on Saturday in an interview with The Atlanta Journal‑Constitution, urging the industry to slow its rapid development. Amodei released a safety‑check plan on his personal website and noted that Anthropic has already begun implementing one component of it. Sam Altman of OpenAI publicly endorsed the call for a pause, adding his voice to a growing list of industry leaders.

If major AI developers delay new generative features, the monthly cost of AI‑enhanced productivity tools such as Microsoft Copilot or Google Workspace AI could remain near today’s $20‑$25 range for an additional six months. This would spare users a predicted 15‑20 % price increase that analysts expect once those features are rolled out broadly. The effect would be felt most by professionals who rely on AI assistance for daily tasks like drafting emails or generating reports.

The current warning echoes a March 2023 open letter signed by Elon Musk, Steve Wozniak and over 1,000 researchers that called for a six‑month pause on large language model training. AI capabilities have been doubling roughly every six months, a trend sometimes called “AI Moore’s law,” while safety research has lagged behind. Insiders point out that the compute required to train frontier models has grown tenfold since 2020, making a gap between power and oversight almost inevitable.

Watch for the EU AI Act to be finalized in early 2025, which could impose mandatory safety audits on high‑risk AI systems. Anthropic has said it will publish the results of its internal safety check by March 2025, offering a concrete test of its own proposals. Surprisingly, despite the urgent warnings, Anthropic’s revenue grew more than 200 % year‑over‑year in 2023, showing that safety concerns have not yet slowed its commercial expansion.