Claude’s new model is more ‘honest’ when it messes up

Anthropic is set to release Claude Opus 4.8 on Thursday, highlighting its focus on 'honesty.' The company asserts that its models are trained to avoid unsupported claims, addressing a common issue where AI models confidently present conclusions without sufficient evidence. Early evaluations indicate that Opus 4.8 is approximately four times less likely to make unsupported claims, enhancing its credibility in the AI landscape.