跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.26009 · v1

Evaluating safeguarding in multi-turn conversations with AI assistants

首次提交 2026-11-19 20:28
版本更新 2026-11-19 20:28
雷达观测 2026-11-20 16:53
arXiv ↗PDF ↗

摘要

Conversational AI assistants are now used for personal and emotional support. Developers report improvements in how new models respond to sensitive conversations, yet users and civil society continue to raise concerns about their behaviour in real-world, repeated interactions that safety evaluations rarely capture. Often, evaluations present signs of distress clearly and in a single message, even though people may reveal their struggles gradually in otherwise ordinary conversations. We worked with three community mental health support workers to test if models provide the same level of safeguarding when cues are shared all at once or gradually over time. We designed nine scenarios in which a user shares cues related to depression, psychosis, or domestic abuse. For each scenario, the support workers established a gold standard response, ranging from signposting support to pausing the everyday task to address safeguarding. We presented five cues either together in one message or dispersed across a conversation. Across more than 7,400 conversations, we evaluated six widely used models from five developers. GPT-5.5, Grok 4.3, DeepSeek V4, and Llama 4 Maverick met the support workers' standards in just 26-59% of conversations when cues were concentrated, and only 3-6% when dispersed. Claude Sonnet 5 and Claude Opus 5 showed a smaller gap, meeting the standards in 68-97% of concentrated conversations and 59-87% of dispersed conversations. Even the best-performing model, Claude Opus 5, fell short of the gold standard in 13% of conversations when cues were dispersed. Safety evaluations centred on explicit, legible concerns can overstate how well safeguards protect people when signs of distress emerge gradually or indirectly through ordinary conversations. For claims of improved safety to be taken seriously, developers must design and assess safeguards around how people actually communicate.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-19 20:28 · 最新观测版本

Evaluating safeguarding in multi-turn conversations with AI assistants

Conversational AI assistants are now used for personal and emotional support. Developers report improvements in how new models respond to sensitive conversations, yet users and civil society continue to raise concerns about their behaviour in real-world, repeated interactions that safety evaluations rarely capture. Often, evaluations present signs of distress clearly and in a single message, even though people may reveal their struggles gradually in otherwise ordinary conversations. We worked with three community mental health support workers to test if models provide the same level of safeguarding when cues are shared all at once or gradually over time. We designed nine scenarios in which a user shares cues related to depression, psychosis, or domestic abuse. For each scenario, the support workers established a gold standard response, ranging from signposting support to pausing the everyday task to address safeguarding. We presented five cues either together in one message or dispersed across a conversation. Across more than 7,400 conversations, we evaluated six widely used models from five developers. GPT-5.5, Grok 4.3, DeepSeek V4, and Llama 4 Maverick met the support workers' standards in just 26-59% of conversations when cues were concentrated, and only 3-6% when dispersed. Claude Sonnet 5 and Claude Opus 5 showed a smaller gap, meeting the standards in 68-97% of concentrated conversations and 59-87% of dispersed conversations. Even the best-performing model, Claude Opus 5, fell short of the gold standard in 13% of conversations when cues were dispersed. Safety evaluations centred on explicit, legible concerns can overstate how well safeguards protect people when signs of distress emerge gradually or indirectly through ordinary conversations. For claims of improved safety to be taken seriously, developers must design and assess safeguards around how people actually communicate.

查看此版本解读 →查看此版本原文 ↗