Anthropic co-founder calls for mandatory third-party AI kill switch

September 14, 2026 at 11:54 PM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

Anthropic co-founder Jack Clark said AI companies may need a mandatory kill switch verifiable by third parties. Clark told the BBC that society might want to pass rules around shutting off AI software if it gets too dangerous. His comments come as US lawmakers push the Kill Switch Act while President Donald Trump rejects slowing AI development.

Key Facts

Jack Clark, one of seven co-founders of Anthropic, said a third-party-verifiable kill switch for AI may need to be mandatory.
Anthropic scientist Evan Hubinger said he personally puts the probability of human extinction from AI at more than 10% within the next decade.
US lawmakers have introduced the Kill Switch Act, which would require companies to have a way to shut down problematic AI tools and give government agencies power to demand a tool be turned off or limited.
US President Donald Trump called the idea of AI destroying humanity a hoax and said there is a conspiracy against AI and data centers that only benefits China.

The Kill Switch Proposal

Jack Clark, one of seven co-founders of Anthropic, told the BBC that a way of shutting off AI software completely if it gets too dangerous was the kind of thing society might want to eventually pass rules around. Clark said most labs, including Anthropic, have different ways of being able to pull the plug, but policymakers may need to enforce having one. He said specifics around kill switch requirements and third-party verification should be part of the larger policy conversation taking place around AI. Clark added that allowing AI to continue as a totally unregulated industry was a bad idea, saying, "We are rolling dice with immense risks."

Anthropic's Safety Debate

Anthropic, formed in 2021 by former OpenAI employees, is at the centre of a debate around AI safety. Last week, a post from an AI researcher who quit Anthropic over concerns AI could wipe out humanity went viral. In response, Anthropic scientist Evan Hubinger said he personally thought the possibility of human extinction from AI was more than 10% within the next decade. Anthropic head Dario Amodei over the weekend called for the pace of AI development to slow and be more closely monitored, without sacrificing commercial advantage.

US Policy Divide

US lawmakers have put forward legislation dubbed the Kill Switch Act that would require companies to have a way to shut down problematic AI tools. The bill would give certain government agencies the power to demand a tool be turned off or limited. US President Donald Trump has rejected the idea of any attempts to slow down AI, saying on social media that AI taking over the world and destroying humanity is a hoax. Trump said in a separate post that there is a sick conspiracy going on against AI and data centers, and the only one happy about it is China.