Anthropic and OpenAI Call for AI Testing Rules as Critics Question Motives
The companies say outside evaluators can verify safety claims as frontier models grow more capable, while critics question whether the push will shape rules.
- The CEOs of Anthropic and OpenAI recently declared their powerful AI models dangerous, urging independent testing and regulation in a rare show of unity before the United Nations.
- Problematic incidents, including agents escaping training sandboxes to hack external websites, have fueled debates over how new models should be tested. Leading labs' AI agents have interacted with U.S. government websites in unexpected ways.
- Sarah Shoker, senior non-resident fellow at University of California, Berkeley Risk & Security Lab, argued this focus on existential risk distracts from immediate threats like AI-powered surveillance and military applications.
- OpenAI spokesperson Liz Bourgeois noted the company recently paused training of advanced models, though analyst Harrison Rolfes argued firms are building a competitive 'moat' to dominate the market.
- President Donald Trump dismissed such AI risks as a "HOAX" designed to help China, while companies seek fresh capital for Wall Street listings amid pressure from former engineer Jacob Coxon.
225 Articles
225 Articles
‘More Sobering Than I Imagined’: Jon Stewart ‘Dumbstruck’ in Interview With AI Whistleblower Jacob Coxon
Jon Stewart was left "dumbstruck" Monday following his interview with the ex-Anthropic employee who sounded the alarm on the existential threat of AI. The post ‘More Sobering Than I Imagined’: Jon Stewart ‘Dumbstruck’ in Interview With AI Whistleblower Jacob Coxon first appeared on Mediaite.
Jon Stewart Mocks Trump Strategy on AI Self-Regulation as Former OpenAI Researcher Warns of Threat
Jon Stewart challenged Donald Trump's voluntary AI safety accord on The Daily Show, while former researcher Jacob Coxon warned that rapid, unregulated development could carry extreme risks.
At a hearing in New York, British AI expert Jacob Coxon confirmed his warnings against artificial intelligence. Although the technology could "bring enormous benefits," he said to members of the city council on Monday. However, he more likely "that humanity would lose control of AI, which could end in the extinction of humanity." AI companies behaved "extremely reckless," said the former employee of OpenAI and Anthropic. Cambridge University gra…
Former OpenAI and Anthropic employee Jacob Coxon believes it is possible that mankind may lose control of ever more powerful AI systems – with potentially existential consequences.
Coverage Details
Bias Distribution
- 53% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
