Stop relying on the response. If someone may be in immediate danger, get human help first; then preserve only the context needed to explain what happened, report it through the service and any responsible organization, and verify consequential claims independently. The right reporting route depends on the service, your location, and whether the tool was provided by an institution.
1. Stop and assess the risk
Do not follow AI-generated instructions that could put someone at risk. Pause the conversation if continuing could increase distress or expose more personal information. Consider whether the response suggests an immediate threat, self-harm, violence, exploitation, or another urgent safeguarding concern. Treat the output as unverified, not as authoritative advice.
As an Amazon Associate I earn from qualifying purchases.
NIST’s voluntary AI Risk Management Framework identifies validity, reliability, safety, security, accountability, privacy, and fairness as characteristics of trustworthy AI. NIST also cautions that addressing those characteristics individually does not guarantee a system is trustworthy. The framework helps manage risk; it does not certify that a particular AI tool is safe. Read NIST’s AI RMF FAQs.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
2. Get human help if there is urgent danger
If someone faces imminent physical danger, contact the relevant local emergency service. For emotional crisis support in the United States, call or text the 988 Suicide & Crisis Lifeline. It is a U.S.-specific resource, not a universal emergency number; use locally appropriate services elsewhere.
#1 Best Overall
If the tool was used in a school, workplace, or other institution, involve a trusted adult, supervisor, safeguarding lead, or other responsible contact as appropriate. Do not treat a chatbot as a substitute for a responsible person, qualified professional, or crisis service. UK Department for Education standards for learner-facing generative AI products say users experiencing distress should be directed to human help, such as teachers, family, peers, or crisis services. Those standards concern educational products in the UK, not every consumer AI service worldwide. See the UK standards.
3. Keep a concise, privacy-conscious record
If it is safe and permitted, note the service or tool, date and time, relevant prompt and response, and enough context to explain why the output concerned you. A screenshot or export may help if the service allows it. Keep the record private: avoid reposting harmful content or including unrelated sensitive personal information.
The UK Department for Education’s standards discuss robust activity logging for covered learner-facing products, including recording prompts and responses. That is a product expectation in that educational context, not a universal duty for every user or service.
4. Report it and escalate where appropriate
Use the AI service’s current in-product report, feedback, or safety channel. Explain plainly what was unsafe and include only the relevant context. Reporting interfaces differ by provider, so consult the service you used rather than assuming one workflow applies everywhere.
If an organization supplied or manages the tool, notify its appropriate safety, safeguarding, or IT contact when warranted. In the UK educational context, the Department for Education standards call for a formal complaints mechanism and relevant supervisor or safeguarding alerts. The appropriate route depends on the organization’s policy and your location.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.5. Verify consequential claims before acting
Check medical, legal, financial, safety, or identity-related claims with qualified people and authoritative sources before making decisions. The Federal Trade Commission’s June 2022 report warned about inaccuracy, bias, classification errors, and failures of context in AI tools used to address online harms. It is a qualitative warning about those tools, not an error-rate measurement for all generative AI systems.
Rank #4
In its June 2022 release about the report, FTC Bureau of Consumer Protection Director Samuel Levine said: “Our report emphasizes that nobody should treat AI as the solution to the spread of harmful online content.” The statement concerns using AI to address harmful online content; it is not a direct claim about every chatbot’s output. Read the FTC release.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
