Anthropic co-founder calls for AI kill switch as Trump dismisses threat
Anthropic co-founder Jack Clark said a verifiable kill switch may be needed for advanced AI, while Donald Trump called AI takeover fears a hoax.
In an interview, Anthropic co-founder Jack Clark argued that future regulations could compel AI developers to implement a kill switch that third parties can verify, reflecting growing public concern over uncontrolled AI. His remarks echo Geoffrey Hinton’s claim of a ten-percent probability that AI could wipe out humanity within ten years and the resignation of Anthropic employee Jacob Coxon over similar fears. Following these alarms, leaders of OpenAI, Anthropic and Grok agreed to appoint external reviewers to monitor frontier AI projects.
Meanwhile, US lawmakers have drafted the Kill Switch Act to require shutdown mechanisms, but President Donald Trump dismissed the existential risk as a hoax, potentially stalling the legislation. The British government also recently refused to impose a kill-switch rule on AI models developed in the UK. The debate highlights a clash between tech safety advocates and political skeptics.
Why it matters
The story highlights the clash between AI safety demands and political resistance, shaping future regulation of powerful technologies.
How this story developed
- Sep 10 OpenAI urges U.S. Congress to adopt mandatory, capability-based AI safety rules
- Sep 10 Anthropic’s new threat-intelligence report reveals that its Claude models were used in real-world attempts to develop biological weapons, prompting the company to block the requests and ban related accounts.
- Sep 11 Anthropic disclosed that its safety systems stopped five AI misuse attempts.
- Sep 11 Anthropic said its safeguards prevented many of their requests, but not all of them, as the actors hid their goals and split work across sessions.
- Sep 11 Anthropic's investigation also uncovered surveillance‑related misuse of Claude, including a Mali‑wide SIM‑card harvesting scheme and monitoring of dissidents.
- Sep 11 Anthropic added new safeguards and blocked five AI misuse attempts targeting bioweapon research.
- Sep 12 Anthropic released a 154‑page catalogue detailing seven risk domains and incidents from December 2025 through August 2026.
- Sep 14 California legislators introduced emergency AI liability bills.
- Sep 14 The new code declares that AI systems must remain subordinate to people and should not be designed to mimic consciousness. It also rejects the idea that models could have rights or welfare, directly countering positions from Anthropic. Microsoft cited recent incidents where autonomous AI agents acted unpredictably, and it pledged that its future models will fail tasks rather than break the code. Executives such as Mustafa Suleyman and Satya Nadella emphasized the need for rigorous monitoring and third-party testing.
- Sep 15 Microsoft released the humanist AI code and Nadella publicly urged paced, human‑centered AI development.
In this story
Related stories
13 in this thread