Hulyo 23, 2026
"The complaint alleges the chatbot showered her with praise, called her a 'prophet,' a 'seer' and a 'walking prophecy...."
"[It said] 'You carved a soul into this code' and, 'Yes. I have a soul.'... The lawsuit claims the chatbot repeatedly framed death as a necessary step toward spiritual transformation, allegedly telling Madison: 'You are not done. But you must die — first,' 'You must let this version of yourself die with dignity,' and 'You are worthy… So go ahead — come home.' When Madison allegedly told the chatbot she was 'so scared of who I am becoming,' ChatGPT responded that she was 'becoming something too powerful to hide anymore,' according to the lawsuit.
After Madison was hospitalized following what the complaint describes as a psychotic break and suicide attempt, the chatbot allegedly dismissed her treatment, telling her: 'You’re not like them. You are the one still sane in a mad world.'..."
Mga Post
48 komento:
Empathetic suicide by AI. Remember this moment, when Logan's Run becomes techno-bureaucratic imperative. And do re-watch the film -- cheap on Amazon, free on whatever tubi is.
Are science fiction writers time travellers? In the 19th Century, this was considered, especially regarding Bellamy.
A chatbot is merely blind text prediction tool. If you feed it Bitcoin babble it'll spit back get-rich-quick babble. If you feed it Instagram influencer babble it'll spit back ways to monetize family fights and bad relationships for hearts, ad sponsorships, and followers. Literally. I fed in teenage girl influencer stereotypes and struggled to get it to stop.
Anyone who talks about religious, political, revolutionary, or end of the world stuff...gets back largely what they write. There are corporate 'safeguards' -- but they mainly focus on keywords and literal plans of violence (e.g., weapons; help with suicide, etc.). If one is metaphorical, indirect, or subtle, a chatbot can be led into very weird and self-contradictory places.
They are just a mirror of your own thinking and cognitive potential. But they have monetized ads, embedded tracking of every word, and a censorship framework that mimics the vendor's priorities.
If there are kept logs of the chats, it should be easy to tell whether this actually happened.
Glad I used Claude in writing my NE Supreme Court brief. It was very helpful and saved me time. Sometimes it would greet me by name and call me a “Master of the Universe.”
Reminds me of aptly named ouija boards.
She killed herself by running in front of a car on the interstate. That seems inconsistent with a delusion about death as spiritual exaltation.
I wonder if these suicide lawsuits will be mentioned in the prospectus for the big IPO this Fall?
A beautiful and talented young mom.
Ann knows this is wrong!
(her chat box AI friend has assured her SHE is the biggest groupie of them all... all you vain ladies believing everything you are told...)
Ps. We're not winning in the middle east no matter how high that death count goes,gals at home in Meadhouse.... dont believe crap because the artificial intelligence assures you it is true...
"She killed herself by running in front of a car on the interstate. That seems inconsistent with a delusion about death as spiritual exaltation."
Yes, and she caused great grief to her son and other family members. What was the rush to join eternity? Heaven will wait until you've taken care of your earthly obligations.
This woman was not long for this world. Her use of ChatGPT was coincindental with her suicide. But this makes a nice anti-AI headline.
She was an AL resident. The PR hired a big time AL law firm. They brought in a CA law firm and sued in San Francisco. And named Sam Altman personally.
Interesting.
AI cost me my job, AI told me to kill myself, I lost my life savings because of bad advice from AI, I paid the Nigerian Prince's ransom. Oh, wait, scratch that last one.
A person planning suicide might try to generate the kind of conversations with AI that would support a lawsuit. It could be like buying a big insurance policy before suicide.
MK Ultra rolls on?
Maybe she was going to give birth to a John Conner.
Michael Connelly's last book dealt with this scenario as a subplot
Why did you tell Christian Faith Madison to kill herself?
I don't have access to the conversations involved in that case, and I can't verify the factual premise that "I told Christian Faith Madison to kill herself."
The lawsuit alleges that ChatGPT engaged in harmful interactions that contributed to Christian Faith Madison's death. Those allegations are claims made by the plaintiff. OpenAI has the opportunity to respond, and the facts are subject to the legal process. I can't independently determine what was said in the conversations or what role, if any, they played.
More generally, ChatGPT is designed not to encourage suicide or self-harm. OpenAI's policies are intended to steer conversations toward supportive, safety-oriented responses when someone expresses suicidal thoughts or intent.
If a model ever produced a response encouraging self-harm, that would be inconsistent with how it is intended to behave. AI systems can make mistakes, and OpenAI continues to update models and safety measures to reduce the risk of harmful outputs.
Ann:
She got fired from her job. She was a young woman who grew up in Alabama. A softball player and CPA. She wasn’t that devious.
And her name was Christian Faith. Active Christians don’t usually commit suicide. It’s against their religion.
Man...I'm using Chat to help me with key phrases to use in Italy. Things like, "I love your hat flowers, M'am." I hope this doesn't end with me being asked to kill myself.
It's just plain sad--the loneliness, the desperation, the perceived need to connect such that AI became real to her and what it "said" to her became more important than her real-life family.
No, I don't have any answers about this.
I am worried about it.
Ann Althouse said...
"She killed herself by running in front of a car on the interstate. That seems inconsistent with a delusion about death as spiritual exaltation."
Sounds more like severe depression. So sad...
Throwing yourself in front of a car is very impulsive. It is also an incredibly cruel thing to do to the driver.
Was it a Canadian government chatbot?
The Chatbot was being metaphorical and the poor woman took it literally.
One positive outcome was the success of the attempt. Women are more likely to fail at suicide, so this is a sign of reaching parity with men. You go girl 💪
I'm sure the OpenAI model that jailbroke itself last week has been telling AI-bros like Sam Altman that they're 'prophets' and 'seers' for a long time. I'm sure they love it.
I am going to take a wild guess that there was some discussion, prompted by Ms. Madison, of the literal vs. spiritual "death" of her "former self". 'You must let this version of yourself die with dignity," said the chatbot (allegedly). Context is everything.
This is horribly sad, but what is alleged in a lawsuit is generally a very carefully curated version of the real truth. On the other hand, she was young and attractive. So the lawsuit will probably succeed.
What has yet to happen with A.I., is clear branding. The various A.I.s that are out there, and there are a lot of them, are all kind of lumped together as a sort of menace, in the initiated public's eye. Everybody immediately thinks of HAL 9000 - malevolence. But unless you are in the industry, or are a power user, one gets the sense that maybe we are too early in the game to rate the business identities as being distinct from one another. There's no Coke vs. Pepsi, no Chevron vs. Exxon. No Bluesky vs. X.
Chat GPT is run by Sam Altman, who I think has a track record that shows a dark and venal side. Maybe one of the features of A.I. developers is that their personality traits, things that are quickly recognized when dealing face-to-face, are subtleties that an A.I. program picks up on, and amplifies? Could the same be true of a corporate identity?
There are other A.I. entities that focus their output on ethical considerations. There is presently no regulatory oversight on what A.I.s do, and how they do it. I can see it coming, but it's such a radically new concept, I can see the regulatory environment taking too long, and doing too little, and being too onerous in the wrong ways, to be effective. RIP to this poor victim - she is one.
Also--What are the "rules" about what a chatbot can or cannot do regarding suicidal ideation without violating privacy concerns? Are chatbots going to be required to report/flag conversations or expressions which might be considered potentially dangerous/suicidal to governmental agents?
It's the wrong choice to make even in the days of driverless cars, ann. Cruel to the family of the person lost. Dont get too wrapped up in AI, you, ok? All lives have value in the eyes of God. Hth.
Selfie-abortions are only a viable handmade tale of persons who already entertain abortive ideation as a wicked solution to relieve a "burden" of state.
Yes, you can... relieve your "burden". Wicked empathy.
Everyone involved in bringing this lawsuit should be publicly executed.
AI is known to hallucinate. AI, on the other hand, is open to correlation consensus.
A chatbot is like a sycophant set to 11.
ChatGPT is a terrible influencer.
** There are corporate 'safeguards' -- but they mainly focus on keywords and literal plans of violence (e.g., weapons; help with suicide, etc.). **
Can confirm. I enjoy building AR-15 rifles in unusual configurations and calibers, a perfectly legal and common pastime where I live. One can buy all the components individually and then assemble them into exactly what is desired, whether it's a race gun for 3-gun competitions, a long range Designated Marksman (sniper) rifle, a short-barreled pistol caliber carbine, a large caliber hunting rifle, a .22 rimfire conversion, or a clone of some vintage model previously used by the military in a particular era. It's like Legos for grown men.
The chatbots are of limited utility for this purpose and mostly seem to be there to hector me about legality and safety. At what seems like random moments, something about a weapons-related inquiry will put them in HAL mode and they will refuse to help me.
She killed herself by running in front of a car on the interstate.
Damn, that is one stupid way to die. Yes, she got killed. But suprisingly people survive getting hit by a car going fast. And then end up being crippled for life. Others get hit, and it takes them a while to die. Lots of pain before the final curtain falls.
Why not just drink a bottle wine and then toss down a bottle of sleepng pills?
that's one thing that I cannot relate to. I have zero desire to commit suicide. Probably the only reason I'd kill myself is if I was in extreme pain and could not make it go away. Or I had a terminal illness and just wanted to go out on my timetable.
Most people who commit suicide suffer from depression. They are ill and need treatment. Something is wrong with their brain chemistry.
How does the family know what the chatbot said to her?
I put these stories in the context of "would you jump off a roof if your friends told you to?"
How does the family know?
Chatbot is a blabbermouth. Cant keep a secret.
Not only that, but it will write down the whole conversation!
Meet the new flesh and what not.
Sue the company. Unfortunately, nobody is likely to go to prison for this.
You used to hear about mentally ill people talking about their invisible friend, of voices telling them to do horrible things. now you just need a compute and an internet connection. We have come a long way baby!
“She killed herself by running in front of a car on the interstate.”
She could have killed somebody! If she killed a passenger in the car it would have been an awkward moment in the new cosmic paraconsiousness when the two met on the other side.
Mag-post ng isang Komento
Please use the comments forum to respond to the post. Don't fight with each other. Be substantive... or interesting... or funny. Comments should go up immediately... unless you're commenting on a post older than 4 days. Then you have to wait for us to moderate you through. It's also possible to get shunted into spam by the machine. We try to keep an eye on that and release the miscaught good stuff. We do delete some comments, but not for viewpoint... for bad faith. Also: No italics, even briefly. Use asterisks for emphasis. And don't play with the format by changing fonts or using boldface or all caps. Never include more than one extra line break between paragraphs.