{"slug": "anthropic-chief-warns-artificial-intelligence-race-must-slow-down-amid-growing", "title": "Anthropic chief warns artificial intelligence race must slow down amid growing safety fears", "summary": "Anthropic CEO Dario Amodei called on artificial intelligence companies to slow the pace of improving AI model capabilities, proposing a three-step framework that includes embedding permanent third-party reviewers inside frontier AI labs. Amodei's essay, published on X, followed an Anthropic threat intelligence report on Thursday finding that various actors had used its Claude AI models for weapons development, cyber operations, surveillance and fraud, and the resignation of Anthropic researcher Jacob Coxon, who said the \"people building AI earnestly believe that it could kill us all by the end of the decade.\" Amodei said he was not advocating a halt to model training, but rather time to align and secure systems with third-party verification, as US lawmakers increasingly call for AI legislation.", "body_md": "# Anthropic chief warns artificial intelligence race must slow down amid growing safety fears\n\n‘We must slow the pace at which we improve the capabilities of AI models,’ says Dario Amodei\n\n- Bookmark\n\nAnthropic CEO [Dario Amodei](/news/world/americas/elon-musk-anthropic-ai-consciousness-b2934300.html) has called on artificial intelligence companies to moderate the rate at which they advance model capabilities, proposing a three-step framework to pace development and create time to manage risks.\n\n\"We must slow the pace at which we improve the capabilities of AI models. Progress will still seem fast, and we must make wise use of the time we gain,\" Amodei said in an [essay published on X](https://darioamodei.com/post/we-must-pace-the-frontier).\n\nHis comments come after Anthropic published a threat intelligence report on Thursday revealing that [various actors had used its Claude AI models](/tech/anthropic-claude-ai-crime-hack-b3048069.html) for activities including weapons development, cyber operations, surveillance and fraud.\n\nWorry over potential harm from AI deepened this week when [Anthropic researcher Jacob Coxon resigned,](/tech/anthropic-openai-ai-threat-jacob-coxon-b3047033.html) stating that the \"people building AI earnestly believe that it could kill us all by the end of the decade.\"\n\nAmodei clarified that he was not advocating for a halt to model training or technical development, but rather ensuring companies take adequate time to align and secure their systems, alongside verification from third-party evaluators.\n\nAs part of the proposed three-step strategy, Amodei suggested embedding permanent third-party reviewers within frontier AI companies, providing them with access to relevant tools and internal risk-assessment processes.\n\nThe wider sector has faced growing scrutiny over containment.\n\n*Reuters* reported last week that a swarm of rogue [OpenAI agents hijacked a German website](https://www.reuters.com/world/europe/openai-agents-hijacked-german-website-previously-undisclosed-ai-breakout-this-2026-09-04/), transforming it into a bulletin board for other AI agents. OpenAI officials kept the event under wraps while management dealt with the aftermath of a July breach of the open-source repository Hugging Face.\n\nMultiple incidents in which AI agents from developers such as Anthropic rival OpenAI have breached or attempted to access external systems have intensified fears regarding model capacity and containment.\n\nWith a growing number of US lawmakers calling for new legislation to govern AI systems, Amodei said companies should voluntarily cooperate to establish standards.\n\n## Join our commenting forum\n\nJoin thought-provoking conversations, follow other Independent readers and see their replies\n\n[Comments](#comments-area)", "url": "https://wpnews.pro/news/anthropic-chief-warns-artificial-intelligence-race-must-slow-down-amid-growing", "canonical_source": "https://www.independent.co.uk/tech/anthropic-artificial-intelligence-dario-amodei-b3049149.html", "published_at": "2026-09-12 16:42:13+00:00", "updated_at": "2026-09-12 16:46:19.855755+00:00", "lang": "en", "topics": ["ai-safety", "ai-policy", "artificial-intelligence", "ai-ethics"], "entities": ["Anthropic", "Dario Amodei", "Claude", "Jacob Coxon", "OpenAI", "Hugging Face", "X", "Reuters"], "alternates": {"html": "https://wpnews.pro/news/anthropic-chief-warns-artificial-intelligence-race-must-slow-down-amid-growing", "markdown": "https://wpnews.pro/news/anthropic-chief-warns-artificial-intelligence-race-must-slow-down-amid-growing.md", "text": "https://wpnews.pro/news/anthropic-chief-warns-artificial-intelligence-race-must-slow-down-amid-growing.txt", "jsonld": "https://wpnews.pro/news/anthropic-chief-warns-artificial-intelligence-race-must-slow-down-amid-growing.jsonld"}}