Anthropic’s Opus 5 Is Better at Resisting Prompt Injection
ID: 3d852a6d-bd57-5bd0-9a32-e3c9bbea02d7
STIX ID: report--3d852a6d-bd57-5bd0-9a32-e3c9bbea02d7
Feed Name: Schneier on Security
Anthropic’s Opus 5 demonstrates improved resistance to prompt injection compared with Opus 4.8 and several non-Claude models on the IPI benchmark—reducing attacker success rates (e.g., to 2.0% within 15 attempts). The post compares robustness across contemporary LLMs, notes prompt injection cannot be fully prevented in general, but highlights measurable progress in specific cases.
Your team is not currently subscribed to this feed. You must subscribe to it in order to see this post.