ChatGPT Health protects the person who signs up for it. It does nothing for the customer, client or colleague whose health details that person pastes in. Only a written policy covers that gap, and it needs one new sentence this month.
On January 7, 2026, OpenAI opened a waitlist for ChatGPT Health, a separate space inside ChatGPT for health questions, medical records and fitness data. The privacy promises are decent. They’re also made to one user about that user’s own information, and your company isn’t a party to them. Think of ChatGPT Health as a good lock on somebody else’s front door. Nothing wrong with the lock. It just isn’t on your building.
Key points
- What did OpenAI promise with ChatGPT Health? OpenAI says Health runs as its own space with separate storage and memories, so health details stay out of ordinary chats.
- Who do those promises protect? The account holder, and only for their own data.
- Does Canadian law cover the gap? Partly, and unevenly. Canada has no standalone AI law, and as we have explained, privacy law carries the main legal penalties for AI use.
- What goes in the staff rule? One sentence, added to whatever AI or confidentiality policy you already have.
What did OpenAI promise with ChatGPT Health?
OpenAI says Health runs as its own space with separate storage and memories, so health details stay out of ordinary chats. It describes added encryption and isolation, optional multi-factor sign-in and a commitment not to train its foundation models on Health conversations. Users can view or delete Health memories and disconnect apps at any time.
The claims are specific, which counts for something. Connectors include Apple Health, Function, MyFitnessPal, Weight Watchers, AllTrails, Instacart and Peloton. Medical records come in through b.well, a U.S. health data connectivity company, and OpenAI limits that part to U.S. users. Access at the start went through a waitlist for a small group of early users.
We haven’t tested any of this. Everything above is OpenAI’s description of its own product, and it was on a waitlist when this column was written. On July 23, 2026, a note on OpenAI’s page described a wider launch to U.S. adults on web and iOS, so details may have moved since.
Who do those promises protect?
The account holder, and only for their own data. The promises govern what OpenAI does with what a user puts into their own Health space. They say nothing about a team lead pasting an employee’s accommodation note into a chat, or a bookkeeper uploading a client’s medical receipt to get a summary. The settings were never aimed at that case.
Two omissions are worth reading closely. OpenAI’s launch text names Free, Go, Plus and Pro users and doesn’t name Business or Enterprise. It also doesn’t mention HIPAA, the U.S. federal health privacy law. We don’t read either as a flaw. A consumer health tool and a regulated health service are different products, and the line between them gets drawn in contracts and statutes, not in product copy.
Picture a case, and this is our illustration, not a reported one. A 40 person benefits consultancy has an adviser with a personal Plus account. She pastes a client’s sick-leave note into Health to draft a reply. OpenAI’s promises apply to her account. The client never agreed to anything with OpenAI, and neither did the firm. If the client asks where that note went, “the employee’s settings” is a thin answer.
Here’s the practical consequence. A feature description is not a data processing agreement. No company can hold up a settings screen in front of a regulator and call it due diligence. Vendor assurances start to count when they sit in a written agreement on a plan you control, and a staff member’s personal account is neither.
Does Canadian law cover the gap?
Partly, and unevenly. Canada has no standalone AI law, and as we have explained, privacy law carries the main legal penalties for AI use. How regulators would treat staff health details typed into a personal ChatGPT account probably varies by province and by who is holding the data.
We aren’t lawyers, and we won’t pretend that question has a tidy answer.
Compare it with a paper file. Nobody would carry a colleague’s medical note home in a bag and call the bag secure because it has a zip. The same instinct applies on a screen.
That uncertainty is the argument for a short written rule over a guess. If the law is unsettled, the firm that can say “we told staff exactly what not to paste, in writing, on this date” is in a far better position than the one relying on a vendor’s settings page.
What goes in the staff rule?
One sentence, added to whatever AI or confidentiality policy you already have. Staff may use AI health tools for their own health on their own accounts, and that’s their business. No customer, client or colleague health information goes into any AI tool unless it’s on a plan the company has bought and reviewed. Anyone unsure asks before pasting, not after.
Name one person who answers those questions, so “ask first” has somewhere to go.
Most firms already have a rule about confidential data, but it rarely names health, and health details turn up in HR files, benefits claims, return-to-work notes and customer service threads. A rule that names them is easier to follow than one that says “sensitive information”. Our guide to privacy safeguards for AI email workflows shows how short such a rule can be.
Then check what the company actually pays for. List the AI plans you’ve bought, such as the ones in our Microsoft Copilot plan and pricing guide, and read whether the vendor’s terms address health data. If you can’t find the answer in ten minutes, that is your answer. Keep the policy to one screen. Nobody follows a rule they can’t remember. Put the sentence in the onboarding pack too, because new hires arrive with their own habits and their own personal accounts, and the first week is when a rule sticks.
Is this a fuss over a consumer feature?
The sceptic says yes. People already ask chatbots about symptoms and lab results, and a dedicated space with separate memory beats a general chat where health details mingle with everything else. Fair, and we agree it’s an improvement for the individual.
Our point is narrower. The improvement helps the person using it for themselves and does nothing for the firm holding someone else’s information. An employee with a personal account and a colleague’s note is exactly the case the product doesn’t cover, and it’s the case that ends up in an HR complaint.
We’d soften this if OpenAI put the same commitments in writing for business plans, or if independent reviewers confirmed the isolation and training claims. Until then, write the sentence and send it to staff. It takes ten minutes, and most of that is deciding who signs off.
Frequently asked questions
Does ChatGPT Health use my health chats to train OpenAI’s models?
OpenAI says it does not use Health conversations to train its foundation models. That is the company’s own statement, so treat it as a claim until independent reviewers check it.
Can Canadians use ChatGPT Health?
At launch OpenAI named users outside the EEA, Switzerland and the UK, which would include Canada. The medical records import was limited to the U.S., so the feature set is narrower here.
Should employees put customer health information into ChatGPT?
No, not on a personal account, and not on any plan your company has not reviewed. Add a single sentence to your AI policy so the rule is written down.
Archive entry dated January 8, 2026, written and fact-checked on October 8, 2026. Sources are linked on the claims they support.
