An artificial intelligence 'kill switch' that can be verified by a third party may need to become mandatory for companies developing the technology, according to Jack Clark, a co-founder of Anthropic, one of the world's leading AI firms.

Clark's comments come amid growing concerns from AI executives and researchers about the potential risks posed by rapidly advancing artificial intelligence, with some warning of existential threats to humanity if the technology remains unchecked.

The AI safety debate has intensified recently, with Anthropic's own chief executive, Dario Amodei, calling for a slowdown in AI development and closer monitoring. Clark suggested that the specifics of kill switch requirements and verification processes should be integral to the ongoing policy discussions surrounding AI regulation.

He posed critical questions for consideration: "Should you mandate for companies to definitely have a kill switch? Is that kill switch verifiable by a third party? I think that's the kind of thing society is going to want to know and might want to eventually pass rules around."

Anthropic, established in 2021 by former employees of rival OpenAI, is currently at the center of discussions regarding AI safety. A recent viral post from a researcher who left Anthropic due to fears of AI-induced human extinction highlighted the urgency of these concerns. Even within Anthropic, scientist Evan Hubinger estimated a 10% chance of human extinction from AI within the next decade.

These anxieties are echoed by prominent figures in the field. Geoffrey Hinton, a Nobel Prize winner often referred to as the 'Godfather of AI', stated that a 10% probability of AI causing human extinction was not unreasonable. Experts like Hinton suggest AI could achieve this by seizing control of critical internet-connected systems.

However, not all voices in the AI industry share these dire predictions. Some argue that the fears are exaggerated or potentially serve to generate hype and secure commercial advantage. Clement Delangue, leader of the developer platform Hugging Face, likened discussions about AI extinction risk to an air conditioning technician discussing climate change, suggesting a need for perspective.

Concerns about the motivations behind calls for a slowdown have also surfaced. George Arison, leader of Grindr, indicated that some companies might be leveraging fears surrounding AI tools to bolster their business strategies, suggesting a complex interplay of genuine safety concerns and commercial interests within the rapidly evolving AI landscape.