Anthropic skirts UK safety review amid rising AI transparency concerns

Anthropic has opted not to submit its latest AI model for testing to the UK's top AI safety institute, raising concerns about the lack of transparency and oversight in the realm of frontier AI. This decision comes amid growing worries in the UK that such reviews may become a US-only responsibility, potentially undermining public trust in the technology.

The issue has been amplified by the Trump administration's refusal to release details of its AI security framework, which has prompted prominent organizations on both the left and right to call for its transparency, warning that a lack of accountability in the governance of critical technology could result in a significant loss of public trust.

In response to these concerns, Anthropic has agreed to provide "wide-ranging access" to an institute that reviewed cybersecurity incidents at OpenAI.

Brief written by urgent.news from Semafor's own syndicated text. Machine-written — may contain errors; check the original before relying on it.