‘Gambling with our lives’: AI researcher quits Anthropic
Anthropic safety executive Evan Hubinger says AI could kill humans.
- Why it matters
- Hubinger's departure signals internal disagreement at a major AI safety lab about the risks the company is taking in its own research and deployment decisions.
- Most affected
- Anthropic researchers and employees — Internal disagreement over AI safety priorities becomes publicly visible.
Users of Anthropic's AI systems — Uncertainty about whether the company's safety measures match the concerns of departing experts. - Likely next
- Whether Hubinger publicly details his specific safety concerns or disputes Anthropic's responses.
Also reported by: TechCrunch, RTE, The Straits Times, South China Morning Post, Semafor, Canberra Times, The Daily Caller
More in Technology
- ‘Could kill us all’: Anthropic insiders say AI has more than 10% chance of ending humanity – soon
- How to Rebuild Your Workforce the Right Way After AI Layoffs
- 'Why would you ruin your career?': OpenAI claims to have cracked one of math's greatest unsolved problems, but mathematicians allege it played dirty
- A.I. Ethicist Tristan Harris Warns Humanity Is Halfway to Full Takeover