Checked fact 509416 Nov 2026Safety and security
Anthropic says it designs multi-turn safety evals that report an appropriate response rate, the share of conversations in which the model behaved appropriately throughout, and that one bad turn results in a failing grade for the whole conversation.
The exact words it rests on
the share of conversations in which the model behaved appropriately throughout; results in a failing grade for the whole conversation
What the source said when we opened it, on 15 Nov 2026.
The source
Building strong safety evaluations (Anthropic Engineering)
Checked
Checked by the notis newsroom on , against the source above.
In the story
Anthropic says single-turn safety evals are saturated at 94% to 98% on its API 16 Nov 2026
Cite this fact
Anyone may quote this address. It does not change; if we correct the story, this page says so.