TECHNOLOGYMay 28 · 17:00 UTCTHE VERGEJay Peters

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

Anthropic is releasing Claude Opus 4.8, which the company claims is more 'honest' by flagging uncertainties and avoiding unsupported claims. Early testers found it significantly less likely than its predecessor to make confident assertions without sufficient evidence.

Nexus surfaces and summarizes. The full story lives at the source.

Mentioned

AnthropicorganizationClaude Opus 4.8topic

Spot something wrong with this article?Report a problem →

Forward this

Related Signal

Adjacent reporting

TECHNOLOGYApr 16 · 16:17 UTC
Anthropic releases Claude Opus 4.7, concedes it trails unreleased Mythos
TECHNOLOGYMay 5 · 17:00 UTC
OpenAI claims ChatGPT’s new default model hallucinates way less
TECHNOLOGYApr 16 · 15:59 UTC
Anthropic releases a new Opus model amid Mythos Preview buzz
TECHNOLOGYMay 13 · 07:29 UTC
AI will soon be capable of telling convincing lies
TECHNOLOGYApr 16 · 17:42 UTC
Claude Opus 4.7 Is Here: Anthropic’s Latest Model Delivers, But It’s a Token Eating Machine
TECHNOLOGYMay 5 · 17:59 UTC
ChatGPT's new default model is more factual and better at personalization