Anthropic co-founder says AI kill switch may need to be mandatory
Jack Clark, a co-founder of Anthropic, told the BBC that society might eventually pass rules requiring third-party verifiable kill switches for AI software.
Call for mandatory safety measures
Clark stated that most labs currently have different ways to pull the plug on dangerous systems. He believes policymakers may need to enforce having a specific shutdown mechanism. This requirement should include verification by an independent third party to ensure reliability.
Broader context of AI risks
These comments arrive as more executives and staff publicly warn that AI could kill all humans. Anthropic head Dario Amodei recently called for slowing the pace of development and closer monitoring. Some observers have questioned the motivations behind these calls to rein in technology.
Policy conversation details
Clark indicated that specifics around kill switch requirements belong in the larger policy conversation taking place now. He asked whether mandates for companies to definitely have a kill switch are necessary. Any action to slow development should happen without sacrificing commercial advantage according to Amodei.
Reported by one outlet
Only one outlet has published this. Nothing here has been checked against a second report, so read it as that outlet's account and follow the link below for the original.
Reported by
1 independent outlet. Headline as published. Links open the original report.