---
title: Anthropic: Claude now rates the safety of each conversation right after it answers ("self-assessment"); catches 94% of high-priority cases a stronger classifier flags, at a 1% false-positive rate | ctxsec
url: https://ctxsec.com/items/anthropic-claude-self-assessment-misuse-detection
description: ""
status: "official (Anthropic Safeguards blog and X post); figures are Anthropic's own evaluation"
published: 2026-10-30
---

> 

[← AI](/ai) · [markdown](/items/anthropic-claude-self-assessment-misuse-detection.md)

official (Anthropic Safeguards blog and X post); figures are Anthropic's own evaluation

2026-10-30

### Anthropic: Claude now rates the safety of each conversation right after it answers ("self-assessment"); catches 94% of high-priority cases a stronger classifier flags, at a 1% false-positive rate

Anthropic's new Safeguards blog describes **self-assessment**: after each response, a short hidden instruction asks Claude for a compact safety assessment (whether the request may be harmful, what kind, a review priority, whether it refused, and a one-line reason). Because the conversation is already in the model's cache, it costs only a few output tokens.

Anthropic says that on Claude Fable 5 it flags 78% of conversations a stronger prompted classifier flags as potentially harmful and 94% of the high-priority ones at a 1% false-positive rate, and that since June it has surfaced large-scale misuse other safeguards missed. It frames this as one layer of defense in depth, citing AI-assisted intrusion campaigns by state, criminal and hacktivist groups. Treat the numbers as vendor-reported.

Sources: [Anthropic Safeguards blog: Scaling misuse detection with self-assessment](https://safeguards.anthropic.com/2026/self-assessment/) · [Anthropic on X: self-assessment announcement (Oct 29)](https://x.com/AnthropicAI/status/2115871889549865429)