Anthropic says Opus 5 is its least prompt-injectable model so far
Simon Willison · rss · 2026-07-25
Simon Willison quotes Boris Cherny saying that, more than the headline eval scores, the most exciting part of Opus 5 is its resistance to prompt injection.
According to the system card, and specifically the section on page 73, Opus 5 is described as Anthropic’s least prompt-injectable model so far, with strong results across PI evals and red-teaming.
More from Safety
- Study: AI reviewing AI misses most reward hacking; best models miss 25%+ of hacks — ronbodkin · 2026-10-26
- Fangcun AI, China's first independent frontier AI safety lab, launches with DeepSeek eval report — JacquesThibs · 2026-10-26
- ESR on the Bugpocalypse: locking down frontier models only hands the game to Chinese open weights — colinmcnamara · 2026-10-26
- OpenCode breach exposed AI API keys of 1.4% of Console users — thione · 2026-10-26
- DHS orders DeepSeek products removed from US civilian agency systems — thione · 2026-10-26
- One hacker breached 100+ company websites using AI, scope unknown — thione · 2026-10-26