The Background: Why the Buzz Around ChatGPT & Advice
Over the past few years, ChatGPT has increasingly been used not just for casual chatting or productivity help but for health queries, legal questions, and other high-stakes matters.
-
For instance, a survey found that about one in six people use ChatGPT monthly for health-related questions.
-
At the same time, AI models aren’t perfect. Mistakes, misleading outputs, and over-reliance have raised concern within healthcare, legal, and regulatory communities.
-
These concerns are driving liability, safety, regulatory, and reputational issues for OpenAI.
In late October 2025, OpenAI updated its usage policy in a way that many interpreted (or misinterpreted) as “ChatGPT will no longer provide health or legal advice”. But the reality is more nuanced.
What the Policy Update Says (and Doesn’t Say)
On 29 October 2025, OpenAI updated its “Usage Policies” to include a clause around high-stakes domains, including legal and medical advice:
“… provision of tailored advice that requires a license, such as legal or medical advice, without appropriate involvement by a licensed professional.”
It also added language about the automation of high-stakes decisions in sensitive areas without human review.
Key points:
-
The policy draws a clearer line between general information, which ChatGPT can provide, and personalised advice, which requires a licensed professional.
-
OpenAI emphasised that this is not a new change to what ChatGPT does, but an update/refinement of its existing guidelines.
-
In reality, the chatbot’s behaviour remains mostly the same: it still answers questions about health, law, etc., but adds disclaimers and pushes professional help when appropriate.
Health & Medical Advice: What ChatGPT Can and Can’t Do
What it can do:
-
Explain health topics, medical conditions in plain language (for example: what “Bell’s palsy” is).
-
Offer general guidance (“you might rest, stay hydrated, speak to your doctor”) rather than diagnosing.
-
Help users prepare for their doctor’s visits by breaking down jargon, summarising options, and explaining typical next steps.
What it can’t/shouldn’t do:
-
Provide a formal diagnosis or prescribe treatment specific to an individual’s case.
-
Replace a licensed medical professional in a serious situation. For example, in a test scenario, ChatGPT recommended calling emergency services when a symptom of facial immobility arose.
-
Guarantee accuracy or be used as a sole basis for a major health decision. The policy emphasises human review.
Legal & Professional Advice: The Same Story (with a Twist)
What ChatGPT can do:
-
Explain legal processes, general law concepts (e.g., employment law principles). For example, it generated a contract template and discussed negligence law in a recent test.
-
Provide general legal information (which is distinct from formal legal advice).
What it can’t/shouldn’t do:
-
Give tailored legal advice (i.e., specific to a person’s unique facts) without a licensed attorney.
-
Act as a substitute for an attorney in high-stakes cases. Although it may draft documents or templates, OpenAI emphasises the need for professional review.
In short, yes, it still does legal-help type tasks, but the policy tries to place the burden and liability properly with professionals.
Why This Matters?
This isn’t just a semantic shift. There are several reasons why this update is important:
-
Liability: If an AI gives faulty medical or legal advice and someone is harmed, who is responsible? OpenAI’s policy update helps clarify coverage.
-
Trust & Safety: Users may assume ChatGPT is the same as a doctor/lawyer. By drawing clearer boundaries, OpenAI seeks to reduce the risk of misuse.
-
Regulation: Governments and regulators are increasingly scrutinising AI in sensitive domains. A clear policy helps prepare for future regulation.
-
User clarity: There was confusion caused by social-media posts claiming ChatGPT would no longer provide health or legal advice. This clarification matters.
Implications for Users
For everyday users:
-
You can continue to ask ChatGPT questions about health and law, but remember it is not a substitute for a professional.
-
Use it as a tool: e.g., to understand terminology, compare options, and prepare for a doctor’s or lawyer’s appointment.
-
Don’t rely on it for final diagnosis, treatment plans, legal rulings, or binding legal documents.
For professionals/businesses:
-
If you are a lawyer, doctor, healthcare provider or firm, be aware of how you or your clients might use ChatGPT and maintain oversight.
-
Ensure any use of AI in professional settings includes human review and licensed professional sign-off.
-
Update internal policies and client disclaimers to reflect the line between “general info” and “licensed advice”.
For OpenAI watchers and regulators:
-
Observe how this policy evolves. Just because behaviour hasn’t changed yet, it doesn’t mean changes are not coming.
-
Monitor how AI-providers handle high-stakes domains and how society regulates them.
The Bottom Line: ChatGPT’s Role Going Forward
In a nutshell:
-
The message from OpenAI is: “We’re clarifying our policy, not removing the feature.” ChatGPT can still answer health and legal questions but not act as your substitute doctor or lawyer.
-
The update draws a firmer boundary between information (OK) and professional advice (requires licensed oversight).
-
For users, the best practice is: use ChatGPT as a companion tool, not as the final authority.
-
As AI becomes more integrated into these sensitive sectors, the expectation of human oversight, professional accountability, and ethical use will only increase.
FAQ
Q1: Does ChatGPT still give medical advice?
Yes, but only in a general, informational sense. It cannot legally or ethically replace a licensed medical professional. The policy update reinforces that.
Q2: Can ChatGPT draft a legal document for me?
Yes, it can draft example documents and explain legal concepts, but those documents should be reviewed by a licensed lawyer before use.
Q3: Has ChatGPT been disabled for health or legal use?
No, the core service remains. What has changed is the clearer communication about what it shouldn’t do (i.e., act as a licensed advisor without oversight).
Q4: What should I do if I need serious medical or legal help?
Use ChatGPT for background or preparation, but consult a licensed professional for diagnosis, treatment, legal advice, or binding decisions.
Q5: Are there risks if I rely on ChatGPT for legal/medical decisions?
Yes. Without professional judgement or oversight, relying solely on ChatGPT could lead to errors, legal liability, or health risks.
Q6: How should businesses use ChatGPT in professional settings?
Incorporate it as a support tool (e.g., summarising case law, explaining medical trends) but ensure licensed professionals oversee and approve any critical output.
Conclusion
The recent policy update by OpenAI is an important reminder of the limits of even advanced AI chatbots like ChatGPT in sensitive domains such as health and law. While the service continues to be a powerful tool for general information and exploration, users and professionals alike should approach it with awareness, caution, and a commitment to human oversight.
—-------------------------------------------------------
Like what you just read? Keep the momentum going with these:
0 Kommentare