Anthropic 为 Claude 推出 self-assessment 安全机制
Anthropic 推出 Claude 自评估机制以识别对话中的潜在滥用
Anthropic 训练 Claude 在每次回复后即时给出对话的初步安全评估，并将该机制称为 self-assessment，用于标记其他安全防线未覆盖的潜在滥用。
Anthropic· @AnthropicAI · X·2026-10-30 02:22· 38 分钟前AI 评分52
AI 导读
Anthropic 推出名为 self-assessment 的安全机制，Claude 在每次回复后即时给出对话的初步安全评估，标记其他安全防线未覆盖的潜在滥用。
正文 · 原文
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/
来源：Anthropic · x.com
来源：RadarAI · aihot.news