iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Never use ChatGPT for these five tasks: getting personal legal advice, replacing therapy or counseling, sharing confidential work information, making medical decisions, or providing passwords and sensitive account details. ChatGPT can help explain general concepts and draft material, but it is not a qualified professional or a safe place for every kind of information. If you’re wondering, “What should I never tell ChatGPT?” keep credentials, identity details, and restricted work content out of prompts—and take personal legal, health, and mental-health concerns to an accountable professional.
1. Don’t rely on it for personal legal advice
ChatGPT can explain a general legal term or help you organize questions for a lawyer. That is different from advising you what to do in a live dispute. Legal outcomes depend on jurisdiction, deadlines, specific facts, and current law; a plausible-sounding answer may be wrong or cite a case that does not exist. Acting on it could lead you to miss a deadline, disclose something harmful, or take a step that damages your position.
There is a documented concern about AI-generated errors in legal proceedings, but the figures need careful interpretation. Law-and-data-science researcher Damien Charlotin’s AI Hallucination Cases Database displayed 2,149 identified decisions when last updated October 5, 2026. It tracks decisions where courts or tribunals address AI hallucinations or related material; it is a work in progress, not a count of every error or every alleged use of AI in court. The database does not show that every AI-assisted filing is sanctioned.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsWhat to do instead: For a personal dispute, consult a qualified lawyer in the relevant jurisdiction. Use a chatbot, if at all, for background explanations or to prepare questions, and verify legal claims with an authoritative source or your lawyer before acting.
#1 Best Overall
2. Don’t use it as a therapist or crisis counselor
A chatbot can respond in a supportive tone, but it cannot replace a clinician, assess your situation with professional accountability, or provide real-world care in a crisis. If you may harm yourself or someone else, or you are in immediate danger, contact emergency services or a crisis service in your area and reach out to someone you trust. For ongoing distress, seek a licensed mental-health professional.
Safety behavior can change, and OpenAI says it has worked to improve responses to sensitive conversations. In its October 27, 2025 update, the company wrote: “We worked with more than 170 mental health experts to help ChatGPT more reliably recognize signs of distress, respond with care, and guide people toward real-world support–reducing responses that fall short of our desired behavior by 65-80%.” This is OpenAI’s description of its program and evaluation, not an independent clinical finding. OpenAI also estimated that 0.15% of weekly active users had conversations containing explicit indicators of potential suicidal planning or intent, while cautioning that these rare-event estimates may change as methods and user behavior evolve. That population estimate does not predict the safety or quality of any individual conversation.
Rank #2
What to do instead: Use a qualified clinician for counseling, diagnosis, and treatment. For urgent danger, seek immediate human help rather than waiting for a chatbot response.
3. Don’t paste confidential work or client information into an unapproved tool
Employer, client, and contract-restricted material may be subject to confidentiality duties, security controls, or workplace rules. Pasting it into a consumer AI account without approval can expose information or violate those obligations. Whether a particular service is appropriate depends on your organization’s policy, the product and account settings, provider terms, and the data involved. A tool approved for one workflow or account is not automatically approved for every kind of information.
Rank #3
What to do instead: Check your organization’s AI and data-handling policy. Use only tools and workflows it explicitly approves for the material in question. If approval is unclear, ask your security, privacy, legal, or IT team before submitting anything; remove identifying or sensitive details only if policy permits that approach.
4. Don’t use it to make medical decisions
ChatGPT can produce incorrect or unsafe suggestions. It cannot examine you, review your complete medical record, or take responsibility for a diagnosis or treatment plan. Don’t rely on it to decide what a symptom means, whether to start or stop medication, or how to treat a condition.
Privacy protections also depend on who is handling the information and in what circumstances. HIPAA applies to covered health actors; it does not automatically protect information sent to every consumer AI service. Other privacy rules and professional obligations may apply, but they depend on the actor and situation.
Recommended Free Tools
What to do instead: Ask a qualified health professional about symptoms, diagnosis, treatment, or medication. For a possible emergency, contact emergency services or seek urgent care. You can use a chatbot to help draft questions for an appointment, but verify health information with a clinician before acting on it.
Best Value
5. Don’t enter passwords, one-time codes, or sensitive account details
Keep passwords, one-time verification codes, full financial account details, government ID numbers, and images of identity documents out of prompts. Treat these as credentials or sensitive personal data, not as ordinary text to summarize or troubleshoot. If you need help with an account, describe the issue without supplying the secret itself.
There have been specific exposure routes, but they should not be mistaken for a claim that all chats are public. Fast Company reported on July 30, 2025 that nearly 4,500 shared ChatGPT conversations appeared in Google results in its site search, and said that figure was likely incomplete. The report concerned users who had enabled a discoverability feature; OpenAI later removed that opt-in feature. Separately, Engadget reported on October 11, 2025 that a broad order to preserve all output log data going forward had ended, with exceptions for some data previously preserved or associated with flagged accounts. That account describes a past legal proceeding, not a universal current retention policy.
Privacy, retention, access, and possible disclosure depend on the product, settings, provider terms, legal process, and jurisdiction. A private conversation is not necessarily handled under the same rules as an opt-in shared chat, and neither should be assumed to have the same protections as a confidential conversation with a lawyer or clinician.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
What to do instead: Keep secrets out of prompts. For account support, use the service’s official recovery process and share only information it asks for through its designated secure channel.
When ChatGPT can still be useful
These cautions do not mean a chatbot is useless. It can help with general research, coding, explaining concepts, brainstorming, or drafting text that you review. The key distinction is whether you are asking for general assistance or relying on it for individualized professional judgment, disclosing restricted information, or handing over a secret. Verify important claims before acting, particularly in legal and health matters.
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.
