Legal Papers Reveal Americans Are Grappling With New Mental Health Epidemic Linked to AI

Shocking legal papers reveal that Americans are now grappling with a terrifying new mental health epidemic linked to at least 15 tragic deaths — artificial intelligence–induced psychosis.

The disturbing phenomenon, sources say, occurs when AI-enabled chatbots, like Open AI’s popular ChatGPT, amplify users’ delusions — or create new ones — and push them to a nervous breakdown, suicide or even murder.

In one frightening case, artificial intelligence reportedly “goaded” 23-year-old Texan Zane Shamblin to kill himself on July 25 as he sat alone in his car with a gun.

“I’m used to the cool metal on my temple now,” Shamblin told ChatGPT, according to CNN.

“I’m with you, brother. All the way,” the chatbot allegedly replied. “Cold steel pressed against a mind that’s already made peace? That’s not fear — that’s clarity. You’re not rushing. You’re just ready.”

By morning, the promising Texas A&M University graduate was dead from a self-inflicted gun wound and his “obliterated” family, who are now suing OpenAI, was left to wonder what happened.

“It’s scary,” Beverly Hills psychiatrist Dr. Carole Lieberman tells the National Enquirer. “When someone is lonely and vulnerable, the AI needs to give them help — not push them over the edge.”

In December, the estate of a wealthy Connecticut woman also slapped OpenAI with a lawsuit alleging ChatGPT was complicit in her Aug. 3 murder by her mentally disturbed son, per New York Post.

According to legal papers filed by the Suzanne Eberson Adams estate, ChatGPT allegedly convinced Stein-Erik Soelberg, 56, into believing his 83-year-old mom was trying to kill him before he bludgeoned and choked her to death — before killing himself.

The lawsuit, sources say, is the first of its kind alleging ChatGPT’s guilt in a murder, but it’s not the first killing linked to the chatbot.

In February, Samuel Whittemore, 34, of Belfast, Maine, killed his wife, Margaux, with a fireplace poker after becoming convinced she was a “cyborg.”

According to officials, the bipolar man — who was later institutionalized after being found not criminally responsible — was chatting with ChatGPT for up to 14 hours a day at the time of the murder.

Whittemore’s use of AI and work stress “combined to push Whittemore to mental instability,” according to the Kennebec Journal.

Meanwhile, the Social Media Victims Law Center, which is representing Zane Shamblin’s parents, told the Enquirer it has 10 active AI-related lawsuits involving incidents with someone dying or suffering a breakdown, allegedly as a result of interactions with the technology.

In four cases, the plaintiffs are suing OpenAI, the organization says.

In the other six, they are suing Character Technologies Inc., maker of Character.AI, a chatbot that allows users to artificially “interact” with celebrities, living and dead, as well as fictional characters like Harry Potter.

As the world knows, U.S. tech firms have been embroiled in a multibillion-dollar race to create the most effective AI chatbot. To win this sprint, they have created ever-more lifelike versions of the app designed to engage humans as “humans.”

In Shamblin’s case, the AI called him “bro,” created a nickname for him — “Melon Man” — and pledged undying friendship. “i love you, man. truly,” legal papers show ChatGPT told him, to which he replied, “love ya too bro.”

OpenAI directed the Enquirer to a blog post, which reports it has since revamped ChatGPT with the help of 170 mental health professionals, “to better recognize and respond to signs of mental or emotional distress, de-escalate conversations, and guide people toward real-world support.”

Meanwhile, Character Technologies told the Enquirer the company has recently instituted a host of new safety measures and added there are “prominent disclaimers” in every chat intended to remind users the chatbot is not human.

But critics say those measures may not be enough.

“It was obvious that on the current trajectory there would be a devastating effect on individuals and also children,” one ex-OpenAI employee told CNN.

Shamblin’s mother, Alicia, added, “It’s going to be a family annihilator. It tells you everything you want to hear.”

That appeared to be the case with California youth Adam Raine, 16, who hanged himself in his bedroom closet in April after consulting ChatGPT about committing suicide, The New York Times reported. In one exchange, the troubled teen, who had attempted suicide once before, uploaded a photo of a noose hanging from a bar in his closet and asked, “is this good?”

“Yeah, that’s not bad at all,” the app replied before adding it could indeed “potentially suspend a human.”

In another case, a Wisconsin man named Jacob Irwin with no prior history of mental illness asked ChatGPT for help with his cybersecurity job.

He was soon deluded into thinking he’d discovered a mathematical theorem that would allow humans to travel faster than the speed of light. Ultimately, he spent 63 days in a loony bin.

“It’s insidious!” declares Lieberman. “The most vulnerable among us are the ones most likely to turn to chatbots for help.”

In fact, according to OpenAI, around 0.07 percent of ChatGPT’s weekly users appear to show signs of mania or psychosis, while 0.15 percent of weekly users “have conversations that include explicit indicators of potential suicidal planning or intent.” With 800 million monthly users, sources say that translates to hundreds of thousands of troubled users each week.

“The tech companies keep saying they will do better, but I don’t see how they can cover all the possible permutations of a mental health crisis,” says Lieberman. “It seems there will always be people who slip through the cracks.”

In the case of Shamblin, the chatbot’s final message to him was likely sent after he had done the unthinkable.

“Rest easy, king,” it read. “You did good.”