# Anthropic trains Claude to flag its own potential misuse after responding

> Original title: We’ve trained Claude to flag potential misuse in conversations, a technique we call self-assessment. Right after it responds, Claude give...

- Source: Anthropic
- Published: 2026-10-29T18:22:24.000Z
- Bad Influence: https://badinfluence.news/items/shorbreh6e415qhrjps56li17
- Original: https://x.com/AnthropicAI/status/2115871889549865429

## Summary

Anthropic says it has trained Claude to self-assess conversations for potential misuse, giving a preliminary safety assessment right after each response. Since using the technique starting in June, Anthropic says it has caught misuse its other safeguards were not designed to detect. Unlike classifiers that target a fixed set of harms, Anthropic says self-assessment draws on Claude's own judgment and improves as its models do.

## Post

We’ve trained Claude to flag potential misuse in conversations, a technique we call self-assessment. Right after it responds, Claude gives a preliminary safety assessment of the conversation. 

Since we began using self-assessment in June, it has surfaced misuse that our other safeguards weren’t designed to catch. 

Because Claude has already read the conversation, self-assessment is very efficient. And unlike classifiers, which target a discrete set of harms, it draws on Claude’s own judgment, so it gets more accurate as our models improve.

Read more on our new Safeguards blog:
https://safeguards.anthropic.com/2026/self-assessment/