AI Experts Demand Independent Safety Checks on Anthropic, OpenAI
Over 100 AI experts signed a public letter calling for independent, transparent safety evaluations of major AI labs including Anthropic and OpenAI.
More than 100 artificial intelligence researchers and experts have signed a public letter demanding that leading AI laboratories, including Anthropic and OpenAI, submit to genuinely independent safety evaluations — a move that signals growing alarm within the scientific community over how frontier AI models are being assessed and governed.
The coalition's core argument is that the current system, in which AI companies largely oversee their own safety testing, creates a fundamental conflict of interest. Experts are calling for evaluators who operate free from financial or institutional ties to the labs themselves, arguing that only truly arm's-length assessments can be trusted to surface genuine risks in powerful foundation models.
Read more Trump Plans Major US Military Presence in Greenland Security Deal →
Transparency is the other major pillar of the letter's demands. Signatories want not only independent oversight but also public disclosure of evaluation methodologies and results — pushing back against a culture of opacity that critics say has allowed safety concerns to be minimized or quietly sidestepped in the race to deploy increasingly capable AI systems.
The pressure on Anthropic and OpenAI comes at a pivotal moment. Both companies have positioned themselves as safety-conscious organizations, yet they face intensifying scrutiny from researchers, policymakers, and civil society groups who question whether internal red-teaming and voluntary commitments are adequate safeguards as AI capabilities accelerate rapidly.
The public letter represents a broader reckoning with who gets to judge whether AI is safe — and whether the industry can be trusted to answer that question on its own terms. Continue reading at US Top News and Analysis.
