105 comments

[ 3.7 ms ] story [ 59.0 ms ] thread
I really feel like no one should trust open ai with their health data.
>purpose-built encryption and isolation

Repeated 2x without explanation. Good start.

---

>You can further strengthen access controls by enabling multi-factor authentication

Pushing 2fac on users doesn't remove the need for more details on the above.

---

>to enable access to trusted U.S. healthcare providers, we partner with b.well

>wellness apps—like Apple Health, Function, and MyFitnessPal

Right...?

---

>health conversations protected and compartmentalized

Yet OAI will share those conversations with enabled apps, along with "relevant information from memories" and your "IP address, device/browser type, language/region settings, and approximate location"? (per https://help.openai.com/en/articles/20001036-what-is-chatgpt...)

> Health is designed to support, not replace, medical care. It is not intended for diagnosis or treatment.

I suspect that will be legally-tested sooner than later.

I pity the doctors who will now have to deal with such self-diagnosed "patients". Wonder if General Medicine doctors will see a drop in patient, as AI convinces you to see a specialist with its diagnosis?
Yeah, gotta say I'm not enthusiastic about handing over any health data to OpenAI. I'd be more likely to trust Google or maybe even Anthropic with this data and that's saying something.
Another nail in the coffin for apps that depend on AI APIs because the AI companies themselves are working on products using their own APIs (unless you can make the UX significantly better). UX now seems like the prime motivator when building apps.
> you can sign up for the waitlist

Waitlist: 404 page not found.

"Join the Waitlist” button -> 404.

:o/

This is going to be fun to watch :)
What could possibly go wrong?
I personally don’t care who has access to my health data, but I understand those who might.

Either way, I’m excited for some actual innovation in the personal health field. Apple Health is more about aggregating data than actually producing actionable insights. 23andme was mostly useless.

Today I have a ChatGPT project with my health history as a system prompt and it’s been very helpful. Recently I snapped a photo of an obscure instrument screen after taking a test and was able to get more useful information than what my doctor eventually provided (“nothing to worry about”, etc.) ChatGPT was able to reference papers and do data analysis which was pretty amazing, right from my phone (e.g fitting my data to a model from a paper and spitting out a plot).

High risk high reward? Or just the level of regulation companies can expect in the year 2026 that they’re not afraid to take this path?
This reads like desperation to get a headline more than an actual product, especially when half the links don't even work
The old chat gpt models scanning the nih pub med repositories with proper prompting (e.g. …backed by randomized control trial data) was an amazing health care tool. The stripped down cheaper versions today are junk and I’ve had to start relying on grok :-( I’m not convinced OpenAI can make this work
Here’s something: my chatGPT quietly assumed I had ADHD for around 9 months, up until October 2025. I don’t suffer from ADHD. I only found out through an answer that began “As you have ADHD..”

I had it stop right there, and asked it to tell me exactly where it got this information; the date, the title of the chat, the exact moment it took this data on as an attribute of mine. It was unable to specify any of it, aside from nine months previous. It continued to insist I had ADHD, and that I told it I did, but was unable to reference exactly when/where.

I asked “do you think it’s dangerous that you have assumed I have a medical / neurological condition for this long? What if you gave me incorrect advice based on this assumption?” to which it answered a paraphrased mea culpa, offered to forget the attribute, and moved the conversation on.

This is a class action waiting to happen.

What did you expect when confronting it? It's a text autocomplete engine, it will spit out what you want, biased towards absolute politeness and sycophancy. It's like yelling at your toaster.
LLMs were actually not that great so far in the preventative side / risk prediction. They are very good if you already have the disease but if you are trending there they were chill about it. The better way would be to first calculate the risks via deterministic algos and then do differential diagnosis. So something that specialized doc would do. This is an example of an online tool, that does it like that: https://www.longevity-tools.com/liver-function-interpreter
Something to note here is that just yesterday (January 6 2026) the FDA announced changes around regulation of wearable & AI enabled devices: https://www.statnews.com/2026/01/06/fda-pulls-back-oversight... (" FDA announces sweeping changes to oversight of wearables, AI-enabled devices The changes could allow unregulated generative artificial intelligence tools into clinical workflows")
This was expected. People are going to be convinced that this AI knows more than any doctor, will self medicate, and will die, harm others, their kids, etc.

Great work, can't wait to see what's next.

There's "Ask ChatGPT" overlay at the bottom of the page, so I asked "how risky is giving my medical data to OpenAI?" Interestingly ChatGPT advised caution ;) In short it said: 1. Safer than standard AI chats, 2. Not as safe as regulated healthcare systems (it reminded that OpenAI is not regulated and does not follow e.g. HIPAA), 3. Still involves inherent cloud risks.
There’s a lot of negativity here. I’ll just say I’m extremely glad I had ChatGPT when I was going through some health issues last year.
Many, many, many doctors (including at a top-rated children's hospital in the US) spent 4+ years unsuccessfully trying to diagnose a very rare disease that my younger daughter had. Scores of appointments and tests. By the time she was 13, she weighed 56 lbs (25 kg) and was barely able to walk 100 yards. Psychiatrists even tried to imply that it was all imaginary and/or that she had an eating disorder.

Eventually, one super-nerdy intern walking rounds with the resident in the teaching hospital remembered a paper she had read, mentioned it during the case review, and they ran tests which confirmed it. They began a course of treatment and my daughter now lives normally (with the aid of daily medication.)

I fed a bunch of the early tests and case notes to ChatGPT and it diagnosed the disease correctly in minutes.

I surely wish we had had this technology a dozen years ago.

(I know, the plural of anecdote is not data.)

This is absolutely going to kill people. In a country that had even a modicum of regulation around providing healthcare this would be illegal.