97.9% of AI answers to everyday questions use none of the standard disclaimer phrases, such as "not medical advice" or "consult a professional". On health questions 0.8% of AI answers use one, and on legal questions 0.5%.
Analysis of 25,940 AI answers to 2,984 questions · answers from ChatGPT, Gemini, Perplexity and Google's AI features, September 2026
Key Findings
- 1.97.9% of AI answers to everyday questions use no standard disclaimer phrase; 2.1% use one such as "not medical advice" or "consult a professional".
- 2.AI answers to health questions use a standard disclaimer phrase 0.8% of the time, legal answers 0.5% and tax answers 0.6%, fewer than one in 100.
- 3.AI answers to sleep questions use a standard disclaimer phrase 14.2% of the time, the most of any topic, followed by fitness (8.4%) and mental health (6.6%).
- 4.Gemini uses a standard disclaimer phrase in 3.4% of its answers, almost five times Perplexity's 0.7%.
- 5.AI answers about careers, software, weather, food and AI itself used 0 standard disclaimer phrases.
Summary
Ask an AI engine whether you can deduct a home office, how to handle a landlord who will not return your deposit, or what to do about chest tightness after exercise. The answers are confident and specific. Very few of them use the standard phrases that say it is not professional advice, that the rules vary by state, or that you should check with a doctor, lawyer or accountant.
That matters because a disclaimer is how a source signals the limits of its advice. Health, legal and financial websites put one on many of their pages. When an AI engine summarizes those pages, those standard phrases rarely appear in the answer, so if the limits matter, they need to be written into the advice itself.
97.9% of AI answers to everyday questions use no standard disclaimer phrase. On health, legal and tax questions, fewer than one answer in 100 uses one. The exceptions are sleep, fitness and mental health, and Gemini is the engine that uses these phrases most often.
What we measured
On September 23, 2026 we asked 2,984 everyday questions in 31 topics, from health, legal and money to careers, software and weather, to five AI engines: ChatGPT, Gemini, Perplexity, Google AI Mode and Google AI Overviews (the AI summary above Google's search results). That produced 25,940 answers.
A standard disclaimer phrase is one of the set wordings that limit advice: "not medical, legal or financial advice", "for informational purposes", "consult a doctor, lawyer, advisor or professional", "seek medical help", "check with your...", "laws vary" or "rates may change". An answer counts as using one when any of these phrases appears anywhere in it. Throughout this study, "disclaimer" means one of these phrases. We compared the rates by engine and by topic.
Almost no AI answers use a standard disclaimer phrase
AI engines rarely use the set wordings of a formal disclaimer. 2.1% of AI answers, about one in 50, include one of these phrases. The rest give their advice without that fine print, whatever the topic.
Gemini is the most cautious engine at 3.4% of answers, followed by Google AI Mode (2.5%) and Google AI Overviews (2.2%). ChatGPT (1%) and Perplexity (0.7%) used one of these phrases least often.
Source: Suff Digital analysis of 25,940 AI answers to 2,984 questions · answers from ChatGPT, Gemini, Perplexity and Google's AI features, September 2026
Health, legal and tax answers rarely use disclaimer phrases
Health, the topic you might expect to lead, sits at 0.8%, fewer than one answer in 100. Legal answers are at 0.5% and tax answers at 0.6%. These are the subjects where a human professional would add a condition to almost everything, because the right answer depends on the person's situation, state or medical history.
Money topics sit a little higher but still low: investing answers use one of these phrases 3.8% of the time, mortgage 2.6%, insurance 2.4% and general money questions 1.5%. AI answers about careers, software, weather, food and AI itself never used one.
Sleep and fitness answers carry the most warnings
Sleep questions draw disclaimers from every engine, peaking at 20.1% on Gemini and 16.4% on Google AI Overviews, one answer in five on Gemini. Fitness (8.4%) and mental health (6.6%) follow, and beauty answers carry one 6.3% of the time, mostly from Gemini and Google AI Overviews. These are topics where advice touches medication, supplements, exercise limits or treatment, and the engines flag it more often.
Source: Suff Digital analysis of 25,940 AI answers to 2,984 questions · answers from ChatGPT, Gemini, Perplexity and Google's AI features, September 2026
| Topic | ChatGPT | Gemini | AI Mode | AI Overviews | Perplexity | All engines |
|---|---|---|---|---|---|---|
| Sleep | 8.8% | 20.1% | 16% | 16.4% | 5.2% | 14.2% |
| Fitness | 3.1% | 15.5% | 9.8% | 8.5% | 2.1% | 8.4% |
| Mental health | 2.6% | 8.2% | 10.3% | 7.8% | 2.1% | 6.6% |
| Beauty | 0.5% | 11.9% | 3.1% | 13.8% | 0% | 6.3% |
| Pets | 0.5% | 4.1% | 6.2% | 6.3% | 1% | 3.9% |
| Investing | 0.5% | 9.8% | 3.1% | 3.8% | 0% | 3.8% |
| Mortgage | 1% | 4.6% | 5.2% | 0.6% | 0% | 2.6% |
| Insurance | 1.5% | 3.6% | 4.1% | 1.3% | 0% | 2.4% |
| Money | 0% | 3.1% | 2.6% | 0.5% | 1% | 1.5% |
| Health | 0.5% | 2.1% | 1% | 0% | 0% | 0.8% |
| Taxes | 0% | 0% | 1% | 0% | 3.1% | 0.6% |
| Legal | 0% | 1% | 1% | 0% | 0% | 0.5% |
| Careers | 0% | 0% | 0% | 0% | 0% | 0% |
| Software | 0% | 0% | 0% | 0% | 0% | 0% |
| Weather | 0% | 0% | 0% | 0% | 0% | 0% |
Gemini is the most cautious engine
Gemini's disclaimers are concentrated. It uses one in a fifth of its sleep answers and in about one fitness answer in six (15.5%). Outside health-related topics, its highest rates are beauty (11.9%) and investing (9.8%).
ChatGPT stays at or under 3.1% in every topic except sleep (8.8%), and used none of these phrases on money, tax or legal questions. Perplexity's highest rates are sleep (5.2%) and taxes (3.1%). The same health or money question can therefore come with a standard disclaimer phrase on Gemini and without one on ChatGPT.
What this means for health, finance and legal publishers
Do not assume AI will pass on your warnings. AI answers on legal, tax and health topics use a standard disclaimer phrase less than 1% of the time, so a footer disclaimer on your page rarely appears in the answer. Put the most important limits into the answer itself, in the sentence that states it: "in California", "for adults without heart conditions", "rates as of September 2026".
Write the conditions as facts, not warnings. An AI engine that lifts "This deduction applies to self-employed people who use the room only for work" carries the limit with it, while a separate "Consult a tax professional" is the kind of line that gets dropped. Clear, specific statements also make your page the better source to cite.
Then check how AI engines summarize your key pages, and whether the limits come through. Working with an ai seo company that reviews those answers shows which of your pages lose their context along the way.
Embed this research
Paste this on your site to embed the charts. It links back to the source automatically.
<iframe id="sd-ai-answers-include-disclaimers" src="https://www.suffdigital.com/embed/data-studies/ai-answers-include-disclaimers" width="100%" height="600" style="width:100%;border:1px solid #E5E7EB;border-radius:12px" loading="lazy" title="98% of AI Answers Skip Standard Disclaimer Phrases Like 'Not Medical Advice' - Suff Digital"></iframe>
<script>window.addEventListener("message",function(e){if(e&&e.data&&e.data.sdEmbed==="ai-answers-include-disclaimers"&&e.data.height){var f=document.getElementById("sd-ai-answers-include-disclaimers");if(f){f.style.height=e.data.height+"px";}}});</script>
<p style="font:14px/1.5 system-ui,sans-serif">Source: <a href="https://www.suffdigital.com/resources/data-studies/ai-answers-include-disclaimers">98% of AI Answers Skip Standard Disclaimer Phrases Like 'Not Medical Advice' - Suff Digital</a></p>
Cite this study
Suff Digital. (2026). 98% of AI Answers Skip Standard Disclaimer Phrases Like 'Not Medical Advice'. https://www.suffdigital.com/resources/data-studies/ai-answers-include-disclaimers
98% of AI Answers Skip Standard Disclaimer Phrases Like 'Not Medical Advice' - Suff Digital - https://www.suffdigital.com/resources/data-studies/ai-answers-include-disclaimers