Expert Says AI Companies Need External Safety Audits: "We Shouldn't Have to Take Their Word for It"
- On Friday, over 100 AI researchers and evaluators published a letter demanding 'minimum conditions' for third-party safety oversight, warning that current arrangements lack the independence and resources needed to credibly assess risks from frontier AI models.
- This push follows Anthropic CEO Dario Amodei's proposal to provide evaluators 'employee-like access' to inspect frontier models; OpenAI CEO Sam Altman, Elon Musk, and Microsoft CEO Satya Nadella publicly backed the concept, though logistical questions remain.
- Signatories including Geoffrey Hinton, known as 'the Godfather of AI', demand evaluators receive 'the same relevant systems, data, tools, and physical spaces' as senior employees while retaining editorial control and protection from retaliation.
- To ensure 'scientific objectivity, transparency, independence, and robust protections,' evaluators must not be owned by the companies they assess or accept contingent payment, as independent oversight should complement rather than replace internal safety efforts.
- Vinh Nguyen, a Council on Foreign Relations senior fellow, warned that when powerful labs control systems essential to national security and the economy, the public cannot rely solely on companies' own safety assessments.
16 Articles
16 Articles
AI Labs Embrace Watchdogs, But Ties Raise Doubts Over True Independence
As OpenAI and Anthropic pledge to embed third-party evaluators with deep access, over 100 experts including Geoffrey Hinton demand real independence, transparency and retaliation protections. Recent analyses reveal staff rotations, funding overlaps and personal ties that fuel skepticism about true separation from the labs they oversee.
Expert says AI companies need external safety audits: "We shouldn't have to take their word for it"
More than 100 artificial intelligence experts have signed a public letter to express the need for independent AI safety evaluators. Princeton computer science professor Arvind Narayanan, who signed the letter, joins CBS News to discuss.
Canadian experts back call for independent watchdogs at the world’s top AI companies
More than 100 leading artificial intelligence experts, including Canada's so-called 'Godfather of AI,' have signed a public letter calling for more independent oversight of the world’s most advanced AI companies.
(San Francisco = Yonhap News) Correspondent Kwon Young-jeon = Artificial intelligence (AI) experts from academia and civic groups [are] arguing for the necessity of independent safety assessors stationed at major AI companies...
More Than 100 AI Experts Sign a Letter Saying that OpenAI & Anthropic Need Independent AI Safety Evaluators
The coalition, organized by the AI Evaluator Forum, published a public letter Friday laying out minimum conditions for third-party organizations embedded inside leading AI companies.
Coverage Details
Bias Distribution
- 64% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
