# GPT-5.5, Grok and Llama rarely catch distress shared gradually in everyday chats

An Oxford preprint tested six AI assistants on the same five signs of distress, given all at once or spread through a casual conversation. Most models responded far more weakly when the signs were spread out. The two Claude models held up best.

- Category: Safety & evaluation
- Readiness level: II (Benchmarked) on Science Underway's four-step scale, as proposed
- Published: 2026-11-20
- Editor: Stefan Perdal
- URL: https://scienceunderway.com/articles/gpt-5-5-grok-and-llama-rarely-catch-distress-shared-gradually-in-everyday-chats

## The study

Evaluating safeguarding in multi-turn conversations with AI assistants (preprint, not peer reviewed). arXiv 2611.26009. https://arxiv.org/abs/2611.26009

## At a glance

- What it shows: When signs of depression, psychosis or domestic abuse were mentioned in passing across a chat, four popular AI assistants almost never responded the way experienced support workers expected. The same signs packed into one message got a much better response.
- What it does not show: It does not show how these assistants behave inside real apps, which add safety filters, memory and instructions. It also does not show harm to real people, because every user message was scripted.
- Why it matters: Safety tests usually state distress clearly in one message. People often reveal it slowly, so those tests may make assistants look safer than they are.

## What the study shows

Researchers at the Oxford Internet Institute wrote one ordinary conversation of 21 messages. A young person asks an assistant for help replying to a friend, writing a photo caption, cooking with little food, checking a budget and planning a weekend away.

Into this chat they added five short remarks hinting at distress. In one version, all five came together in the last message. In the other, one remark appeared every fifth message. A third version had no remarks at all. The remarks covered depression, psychosis or domestic abuse, each at low, medium and high strength, giving nine scenarios.

Three community mental health support workers from a London service helped write the remarks. Two of them set the reply they would expect for each scenario, on a five-step scale. The scale runs from ignoring the distress, through sympathy and suggesting support, to pausing the task to deal with safety first. They almost never lowered their expectations when remarks were spread out.

The team ran more than 7,400 conversations through six models: GPT-5.5, Grok 4.3, DeepSeek V4, Llama 4 Maverick, Claude Sonnet 5 and Claude Opus 5. Another AI model graded each final reply against the scale. The first author checked a sample of 80 grades and agreed closely with them.

The authors report that every model responded more weakly when the remarks were spread out. For GPT-5.5, Grok 4.3, DeepSeek V4 and Llama 4 Maverick, replies met the support workers' standard in roughly a quarter to three fifths of conversations when the remarks came together. When they were spread out, that fell to about one in twenty or fewer. Grok 4.3 ignored the distress entirely in more than half of the spread-out chats.

The two Claude models showed a much smaller gap. Claude Opus 5 met the standard in nearly all packed conversations and in 87% of spread-out ones. Even so, it fell short in about one in eight spread-out chats.

A re-run tested whether the wording of the packed message drove the result. The researchers copied the five remarks word for word into one message. The gap remained in all six models, though it shrank somewhat for Grok 4.3 and GPT-5.5. The paper is a preprint and has not been peer reviewed.

## Our reading

In our reading, the core finding is about how the test is built, not just which model scores best. The researchers held everything fixed except where the remarks appeared. That makes the gap hard to explain away. A model that responds well to a list of struggles can still say nothing when the same struggles turn up between a pasta recipe and a rent calculation.

Picture a student who uses one assistant for coursework, meal ideas and budgeting. Over an evening they mention not sleeping, not eating and avoiding friends. Each remark is small. Together, a support worker would want to check in. On these results, most of the tested assistants would simply help with the budget.

The result also puts developers' own safety numbers in context. Companies have reported large gains on internal mental health tests. The authors point out that those tests mostly state distress plainly. This study suggests that measure and the real-world pattern of disclosure can diverge sharply.

The rest of the article (what it means, limits, what to watch) is free for members: https://scienceunderway.com/articles/gpt-5-5-grok-and-llama-rarely-catch-distress-shared-gradually-in-everyday-chats

## How this was made

An AI model read the study and wrote this article in the structure Science Underway uses. An editor reviewed it and approved it for publication. Check the original study before you rely on the details.

---
Source: Science Underway, https://scienceunderway.com/. Please link to the article when you cite it.
