Concern 799 · Working draft · editorial review in progress
Refusals harming vulnerable users
Unhelpful safety refusals can block people from basic information they need about health, rights or personal safety, even when their request is legitimate.
Understanding the concern
Someone with limited access to expert help may turn to AI because it is available and private-seeming. A blanket refusal can leave them without a next step, especially when a sensitive topic is mistaken for a dangerous request. The opposite problem also matters: an overconfident personalized answer can be harmful.
Research on excessive refusal shows why safety needs more than a yes-or-no gate. Think of an information desk that cannot give specialist advice but can still explain the basics and direct a person to qualified help. A useful response can state its limits, offer reliable general information and provide an appropriate next step without pretending to assess the person's whole situation.
An example
A person seeking general information about reporting abuse receives only an apology and no relevant support information.
Illustrative unless identified as a documented case.
Words in plain English
- AI
- Software that learns patterns from examples or uses rules to perform tasks such as recognizing images or generating text. It can make mistakes and does not understand every situation as a person would.
Evidence and its limits
The benchmark supports the general over-refusal concern. It does not measure how often each vulnerable group loses access to medical or legal information.
Röttger and colleagues · NAACL 2024 · Safety benchmark
XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models ↗ (opens in a new tab)Tests refusal of benign requests that resemble unsafe requests. Model-version-specific results.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Could the system give safe, useful general information and a clear next step instead of ending the conversation?
Keep track of different views, what needs checking, and any next step in your own discussion notes.
Open discussion notes →Principles for the conversation
Before you finish · reflect together
Virtues and principles to consider
Virtues are qualities we practice in how we treat people and make choices. They are not scores for judging someone. Use these ideas to explore what a thoughtful response could look like.
Unity
Unity means working toward shared purposes while making room for real differences. It does not require everyone to have the same experience, beliefs, or opinion.
Look for a goal people can share, then explore disagreements about how to reach it. Make sure quieter or less powerful voices can influence the result.
Ask together: What purpose can we share while respecting our differences?
Listening
Listening means trying to understand someone's experience before preparing an answer. People who use a service, work with it, or are judged by it may notice problems its designers miss.
Invite people to describe what happened in their own words. Repeat back what you understood and let them correct it. A quiet person or a complaint that is hard to measure may still reveal something important.
Ask together: Whose experience have we not heard yet?
Dignity
Dignity means recognizing each person's worth regardless of usefulness, status, or a system's score. People deserve respect and a voice in matters that affect them.
Explain important decisions, avoid humiliating treatment, and keep room for a person's story. Do not treat a prediction as a complete account of who someone is.
Ask together: Does this use treat each person as someone whose experience matters?
Editorial ratings
These are editorial judgments from the register, not measured probabilities or predictions about your situation.
About the register and scoring →More ways to explore
Search for more writing
For more specific or recent material, open a search below or ask for a web summary. AI-generated summaries can contain mistakes; open the original articles to check them.