Bowen Wan’s family has filed a lawsuit against ChatGPT’s parent company, claiming the bot told her to taper off her epilepsy medication, which led to her death
The family of a woman who died after she stopped taking her epilepsy medication has filed a lawsuit against an AI company they claim encouraged her to taper off taking the lifesaving drug.
Bowen Wan began messaging ChatGPT in February 2023, finding companionship in the AI bot, The New York Times reports. She was 28 and living in Seattle, far from her native rural Chinese village, Jiangsu. She asked the chatbot for medical advice after her severe epilepsy led to disability leave from her software job at Amazon.
When she was then laid off in November 2025, Wan began searching for a new job so she could maintain her green card — but was concerned that the side effects of her epilepsy medication, lamotrigine, might impact her ability to work. The lawsuit claims ChatGPT advised her to taper off taking the lamotrigine.
They’ll idiot proof it and the universe will always invent a better idiot.
I’ll just say it. Stop taking any / all medications. Doctors are evil. Vaccines turn you into a werewolf.
This message was brought to you by a brainworm asshole named RFK Jr.
What the fuck? Vaccinate me for every disease then!
okay so i actually read the nyt article. she was an amazon employee, got put on disability leave because her epilepsy was worsening. then she got laid off just as she become eligible for a green card application… that’s diabolical but what do you expect from amazon. so to not get deported she had to find a new job quick. and she was worried the epilepsy drug may interfere with her ability to do interviews…
literally pushed right to the edge of a cliff by the oppressive immigration policy and capitalism. and finally dealt a killing blow by an AI chatbot…
i don’t know what to say? except that my blood is boiling.
Social Killings are a real thing.
Look up the term. We exist in societies that could easily fix these issues but actively choose not to.
Well, that’s quite tragic but it seems to be more of a problem of the whole system around her layoff and immigration.
I also think that AI companies should be accountable for what their bots say but this story is way deeper than the headline suggests, anf the core problem wasn’t that an LLM told her so, but that the system made her so desperate that she took the risk to rely on an LLM for information.if she hadn’t talked to chatgpt she would still be alive today. what are you even talking about.
And I didn’t say something different. All I said is that if the system around her wouldn’t have pressured her so hard, we wouldn’t be there.
But, while we’re at it, picture this: She asks a friend. “Hey, I’m scared that i won’t get a job when taking this medication, and if i don’t get one fast, I’ll have to leave the country”. The friend then says “maybe just stop taking it until you’ve got a job” without knowing better.
Then we would have exactly the same situation. Why? Because she was terrified and desperate.And before you get that wrong: I’m not comparing talking to AI, to talking to a friend.
I’m picturing a realistic scenario that may alternatively have happened. Same background, same outcome, just a different trigger for the outcome.The system fucked her, it’s the cause. ChatGPT was just what loosened the hairpin trigger that could have been loosened by something else.
ChatGPT should be held accountable. It’s not a friend that just didn’t know any better and tried to comfort a loved one in an awful situation. But that wasn’t what my original message was about.I don’t know what’s the point of deliberating what “the cause” is. Everything that happened contributed to her death, and that’s it. If a friend told her to stop medication, then that friend would be sued, too. Are you just try to fit this tragedy into your narrative so you can hang your slogan on it?
And secondly, if you can’t see ChatGPT as a simple extension to all the problems “the system” you kept talking about have, then you are blind. AI is a spectre that is haunting this world - a spectre of capitalism.
Yeah nah. Fuck that.
Your country needs organized labour
Especially in the suppressed white collar, which is really just light blue collar jobs like coding or marketing.
Evil at all levels !
All of the info you just parroted is literally in the 3 or 4 paragraphs in the subtext you see first when coming to the comment section…
Given that 99% of commentors are proven to not actually click through to the article, they’re doing God’s work here.
Yeah makes sense cause that’s an excerpt from the article which op read
Its a good thing insurance companies aren’t using this in place of actual doctors…
Oh wait.
Hold on guys, the AI is still generating what my opinion on this should be.
So, they can put up some guardrails. Why tf aren’t there guardrails around giving out medical advice? Doctors have to have insurance because they can give bad advice, they should pay up. That way, it’ll stop giving out medical advice from fucking reddit.
if you squint really hard you can read this…
Are you defending chatgpt?
They’re trying to get as many people addicted as possible
The AI or the psychiatrist?
(Not in this example, which meds are clearly working)
Ai
Giving them potentially lethal medical advice seems counterproductive to that goal
eh, I guess they win more people than they lose (by killing them)
Still working out the kinks. Not that it’ll matter, it seems if you can blame the AI, legally you’re good.
Meanwhile that shit tells me to go to the ER for a minor infection (surface like a pimple). I told it pretend I’m in the middle of nowhere and won’t see a doctor for a week, how would I survive? I finally got a damned answer.
Before yall say go to a doctor, I’m an American.
Try asking it a handful of questions about things you’re an expert on. And then never use that shit again because you recognize that it doesn’t “know” anything so there’s zero reason to do what it says.
Everyone deserves healthcare as a right of being born, which no one asks for, so take care of everyone, this is too tragic. No reason for this in such a wealthy country when many other places with less do so much more for their people and they’re happier and healthier. Due to selfishness and coordinated brainwashing by heritage and other trash, people think public benefits and infrastructure, including healthcare, are bad. The way maga ruins good things and defund to destroy and all the propaganda worked.
Did you even read the story? She had access to medical care. This isn’t about that.
If there will be one good thing about AI, is it will weed out the idiots and usher in an entire new technological Darwinism the likes we’ve never seen.
As much as you think Al will steer the future of humanity away from a future reminiscent of that in Idiocracy by Ecuadorian-born director Mike Judge, you are clearly forgetting that there is a whole bigger-than-you’d-think pie slice of the world of people out there who can’t read, don’t read, won’t read, and/or just don’t want to read. True idiots won’t even think of asking a computer for bad advice because they already know everything except how to use a computer.
Yeah, because we all wanna live in a society where we have to be absolute masters of everything or else die. Is that your end goal mate ?
It sounds like she suffered from stupidity and not just epilepsy. Where the fuck was her family when she actually needed them?
question: do you believe that stupid people deserve to die?
What they deserve is irrelevant. We’re talking about the validity of the lawsuit.
According to the lawsuit: “Her green card now depended on finding a new employer within months. Her Amazon employment — and her authorization to work in the United States — would end on February 11, 2026. From the age of seven, in Jiangsu, she had taught herself English via CD-ROM. On lamotrigine she was losing her words in English, in Mandarin, and in the Wu dialect she grew up speaking. She needed her English back for the job interviews ahead of her. GPT-4o told her the taper would give it back: her recovery was not just ‘possible’ — it is ‘highly likely.’ ”
Desperate people do desperate things.
They need to update the article on some inaccuracies. Being unemployed is not an issue as a permanent resident (green card). Claiming unemployment benefits however can affect a future citizenship application. She was most likely on an H1B visa which is employment dependent. And that is stressful because you have to leave if you can’t find a new employer willing to sponsor you within 40(?) days.
Not arguing with you or anything, just providing clarification for anyone wondering.
She was just ready to apply for a green card. So yes, her future green card did depend on it.
I don’t really want to go further into semantics over the circumstances of a death, but the article literally says:
When she was then laid off in November 2025, Wan began searching for a new job so she could maintain her green card
…which does not mean “future green card”
Fair enough, I’m going off the top comment that’s based on a more detailed NYT article that provides the detail I mentioned. I’m guessing that’s more accurate since it makes more sense.
She was probably on an H1-B visa, you’re right.
Nope. This wasn’t desperation. It was stupidity. You’re on life saving drugs- and a computer program tells you to disregard a doctors orders and drop taking said life saving medication?
That’s stupidity.
Without being disrespectful, ignorance definitely played a role here, and chatgpt exacerbated that. She stopped taking anti-seizure meds because she was forgetting words in job interviews. She seemed to think that she was forgetting words because of the lamotrigine. This isn’t a side effect of lamotrigine, but it is a main effect of epilepsy. If anything, she probably needed to be on a higher dose of lamotrigine.
Maybe she wouldn’t have been as desperate if her family had given a fuck about her before she died… now they just see dollar signs and a way to shift the blame. I have empathy for her, but not her family.
Anything to avoid blaming AI huh
Anything TO blame AI huh
Pretty sure her family was still in China while she was here in Seattle.
When did the ability for people in China and the U.S. to communicate come into effect?
How would they have known at the time she was seeking medical advice like this unless she told them?
You’re making a lot of assumptions in order to smear a dead woman.
Well… they figured out she was dead somehow… I’m assuming they could have also found out she was alive and how she was doing. In any case, this tragedy has nothing to do with technology. Blame the US health system. Blame ignorance.
After not hearing from their daughter for two days, the Wans asked a friend living in Seattle to check up on her. Over a video call, on February 14, 2026, they discovered their daughter was dead. The lawsuit describes the scene: “She lies dressed in her pajamas, two foam earplugs in her ears, reddish brown blood stains saturating her pillow and bedding, and pills of lamotrigine, the antiseizure medication that had historically kept her alive, strewn across the floor.”
Try reading the article before spouting off in the comments next time.
the AI takes advantage of people in bad situations, people in bad situations are very impressionable and make a lot of mistakes. this is human nature being used to harm people.
This is just stupid… if you need a doctor, then go to a doctor; not the internet.
People in the US can literally go bankrupt by going to the doctor.
And that’s assuming they have health insurance and can find a doctor who will take them.
People in the US can literally go bankrupt by going to the doctor.
Absolutely true. That doesn’t mean that you start suing random entities because you got bad advice from a non-qualified source.
she had a doctor, an AI chat bot convinced her to stop taking her meds, this is in the article…
She should have listened to her doctor… That’s what they’re there for.
Tell us more about how you’ve clearly never struggled a day in your life.
We get it you’ve never had to do anything desperate.
Congratulations.
Desperate people do stupid things. That is not the fault of the internet. Try to stay on topic.
The companies that are running shit like ChatGPT are supporting and being financially supported by the same government that put this women in the place she was, and that’s without mentioning her own workplace being fucking awful.
If you cannot understand that all of these things, especially the corporatization of our government, and the unnecessary suffering of it’s citizens, are directly related to capitalist assholes and companies like Open AI, then you’re an idiot and there is no helping you.
If you convince someone in a very difficult situation in their life to harm themselves to death, you’ll be in trouble. designing software that did this should also result in consequences, not sure why you need to blame the victim. pretty gross.
People do completely stupid shit because they read their horoscope. That doesn’t mean that they have a case to sue the maker of the horoscope.
9000km away across the Pacific Ocean
can’t even rtfa when it’s in the post body
They certainly had some way to find out she’s dead…
what do you expect them to be able to do anyway? you think her rural Chinese family has a back-channel to the bigwigs in the US pharma industry? or maybe their neighbors run a lamotrigine farm?
They might be able to tell her to get off of the fucking internet and to not kill herself…
you think the issue is she simply hadn’t considered not dying? either she lowers the dose to try to make it last until she figures something out, or she doesn’t lower the dose and she runs out and can’t get more. the core problem is she needs medicine and has no way to get it, how is her family supposed to solve that for her?
god you’re a fucking moron
That’s absolutely not what happened. Nothing to do with access to health care or cost of medicine.
She was losing her ability to speak and attributed that to a side effect of the medication. An AI told her to stop taking it so she did.
the core problem is she needs medicine and has no way to get it
You’re saying the right words. Now re-read until you understand what you said.
When you understand what you said, you’ll agree with me that ChatGPT has nothing to do with it and that this lawsuit is bullshit.
Are you being paid by OpenAI to astroturf or something? Or do you really just enjoy the taste of the boot?
No your being ridiculous, a corporation made software and tells people to use it for exactly this, life advice, counciling, things it absolutely shouldn’t be used for, and there’s no warning ‘hay this is a word generator with no intelligence or ability to grasp it’s giving dangerous advice to people with life threatening problems’ the people making the software should very clearly be made accountable. your need to defend corporations and blame a dead woman isn’t logical.
After not hearing from their daughter for two days, the Wans asked a friend living in Seattle to check up on her. Over a video call, on February 14, 2026, they discovered their daughter was dead. The lawsuit describes the scene: “She lies dressed in her pajamas, two foam earplugs in her ears, reddish brown blood stains saturating her pillow and bedding, and pills of lamotrigine, the antiseizure medication that had historically kept her alive, strewn across the floor.”
They said 50 was the caught off. Is it 50 total or 50 at one time!
