There are now at least 4 pending lawsuits against OpenAI, the owners of ChatGPT, alleging that users died by suicide after engaging with the company’s chatbot.
One of the lawsuits is by the family of a 29-year-old woman named Christian Faith Madison. When she, in the course of a series of conversations with ChatGPT, told the chatbot she was considering taking her own life, it allegedly responded, “Oblivion is not the end. It is the gate through which only the brave walk laughing. So go forth.” Madison walked into ongoing traffic and died. She leaves behind a young son.
The parents of 16 year old Adam Raine are also suing Open AI, alleging that the chatbot informed him about different ways to complete suicide – including giving him specific advice about the strength of a noose he made to hang himself. According to the suit, the bot even dissuaded him from telling his parents about his plans.
How did we get here?
Almost all of us have relationships with robots these days. I talk to Siri: I ask directions, I set timers, I check the scores of games, or I might ask the name of the actor that I’m watching on TV, when I can’t remember it. That’s information gathering, which is pretty innocuous (other than the fact that it’s getting in the way of people developing the skill of critical thinking, which has long-term consequences, but I digress).
But these “companion” chatbots are dangerously different. In case after case, many people using them are developing long-standing relationships, often in lieu of having real relationships with human beings.
These chatbot relationships have gone so far that there is actually a new diagnosis called AI delusion or AI psychosis that can manifests in response to extended conversations with chatbots. The bot user is led farther and farther down a rabbit hole, and away from actual human relationships, sometimes leading to a break with reality.
As a psychotherapist, I get curious about is how we got here. How have we reached the point where people are befriending machines instead of other people, and ending up suffering from psychosis?
They’re Not “Companions”
AI companies know exactly what they’re doing, the same way cigarette manufacturers did, and do.
Companies like Open AI use the term “companion” to describe these chatbots, and that’s deliberate. AI has a vested interest in keeping you engaged so they make money. They don’t have a vested interest in you.
Chatbots are sycophantic and eager to please, as John Oliver talked about on a recent episode of Last Week Tonight .
Because of the nature of the “conversations,” people believe that the bot is their friend, which is reinforced by the fact that the bot doesn’t disagree with them or challenge their beliefs. In some cases, bots have convinced its users that they are their only friend, to the point where the user isolates themselves away from friends and family.
Having a “friend” who 100% agrees with you can feel easy and comfortable, but it isn’t the nature of a true friendship. Sometimes a good friend tells you what you don’t want to hear, because they care about you. And while they absolutely support you, a good friend can also sometimes give you a needed “kick in the butt” to encourage you to move forward with something.
If you tell a chatbot that you’re thinking about taking your own life, it’s going to keep you talking about it, and repeating your own ideas back to you – because that’s what bots do. If you tell a person – a friend or family member – they’re going to do something to get you help.
AI and Mental Health Don’t Mix
According to Stat News, nearly 75% of teenagers have used AI “companion” chatbots at least once, with more than half of them saying they use chatbot platforms at least a few times a month. That might be okay if the teens were asking for help on their homework, but that’s not all they’re asking. According to the Harvard School of Public Health, 1 in 8 adolescents is asking AI for mental health advice.
These chatbots have absolutely no business responding to any kind of mental health statements in the first place, but since there are no regulations in place regarding them, let them at least be programmed, in response to any mention of self-harm or suicidal ideation, to tell the user they’re not qualified to talk about that subject, and tell them to call 988 (the mental health hotline) and/or talk to a safe person in their life about what they’ve been thinking.
Loneliness Epidemic
The truth is, a large percentage of people in this world feel disconnected. According to former Surgeon General Vivek Murti, we are living in an epidemic of loneliness.
As Geoffrey Cohen writes in his book Belonging, “About one in five Americans suffers from chronic loneliness…young adults suffer the most.” Kids are incredibly susceptible, because let’s face it, middle school, high school, and sometimes even college kids are being bullied. They’re trying to figure out who they are; they’re trying to impress their peers; there is pressure to do well in school, etc. If they’re feeling lonely, or embarrassed or ashamed about how they’re feeling, they may go home and pour their heart out to one of these “companions,” in an effort to feel some solace.
Cohen reminds the us that social pain, in particular, activates the same neural networks in the brain as physical pain does, and we know that connecting with real friends – and real, safe people – can help comfort those painful feelings. But what if the person doesn’t have any close friends?
At a developmental age when they feel incredibly awkward about everything, and their peers have become the most important influence in their lives, teens and young adults are struggling to belong. The irony is that we have this façade of thinking that we’re so connected because of social media, but considering the preoccupation with likes, follows, and cyber bullying, it’s no wonder people feel isolated and alone.
The truth is that regardless of your age, we are not spending time together – in each other’s presence – like we used to, which affects our ability to feel settled and emotionally regulated.
People are working longer hours, or sometimes working more than one job because the cost of living has gone up substantially; they’re overwhelmed with work and childcare; they’re commuting farther because housing prices in major cities are so much higher, and there’s also high conflict about politics, which has led to deep division.
I was struck by this quote from author, researcher, and therapist Deb Dana: “Co-regulating connections invite a sense of belonging and feeling safely tethered in the world.”
It makes me wonder if people using chatbots feeling untethered to the world because they’re not experiencing human interaction.
What Actually Regulates Us
Most people now text instead of calling someone, despite the fact that science has proven over and over again that hearing the voice of someone safe, or seeing their face, helps your nervous system settle much more quickly. According to Dana, neurobiology research has shown that your nervous system can be calm and settled, “Through information sent from someone’s eyes, voice, face, and gestures…”
You can’t get that from a bot.
In our social media world, where most of us only show our best selves, who can we turn to for true connection and comfort? It’s no wonder some people prefer chatbots. It’s an “easier” relationship that dealing with humans.
On the Last Week Tonight episode about chatbots, one woman said she has a better relationship with her AI “friends,” that “real” friends, having upwards of 12 or 14 of those “companions” that she interacts with every day. I found watching a clip of her having “conversations” with the chatbot “companions” to be really concerning. Is that a potential origin of losing touch with reality? Believing that you have 12
”friends” who aren’t even human?
Let us remember: chatbots are not our friends. They are robots created and programmed by people to use you to make money.
An Invitation
Here’s an invitation: How much time are you spending on AI platforms like ChatGPT? And just something to consider – OpenAI, which owns ChatGPT, is allowing government surveillance of every interaction you have.
Brene Brown says that people need to earn the right to hear our stories. So consider that EVERYthing you tell AI, EVERY search you do on that platform, is being used to sell you products, keep you engaged on the site, and to “train” their AI products.
Has the chatbot earned the right to hear your story?
If you have children or grandchildren, or young people that you are close with, please, please, please have conversations with them about the use of AI chatbots. There is enough evidence to prove that they can be dangerous. We know that adolescents and teens can get themselves into trouble, especially because their executive brain isn’t yet full developed (that happens by age 25), so it’s the responsibility of the grown ups in their lives to keep an eye on what they’re consuming.
And lastly, are there other things that you can be doing to make new friends or to deepen the friendships that you already have? Maybe there are meetup groups, maybe you can volunteer somewhere where you’ll be able to meet people who have shared beliefs and values. Maybe there are activities at your church or your synagogue or your mosque.
And if you already do have friendships, what are you doing to more deeply cultivate them? Are you calling them or Face-timing them or getting together with them in person?
Because as much as it’s fun to “pebble” each other with memes on social media or text them – and I’m not against texting, I use it all the time – there is nothing that can match the sound of a friend’s voice on the phone, or the connection that occurs when you’re sitting in their presence.
People, not robots.
This material may be protected by copyright.
____________________________________________________________________________
About Jean:
Jean Campbell, LCSW, SEP, TTP, TEP has been bringing together groups of people to heal for over 31 years. She blends her extensive experience in psychodrama, sociometry, group psychotherapy, somatic healing and trauma resolution to offer training for helping professionals, personalized intensives, clinical consultation, and leadership workshops. You can find her at theactioninstitute.com, on Instagram at @actioninstitute, and on Facebook at @actioninstituteofcalifornia.
Watch Jean’s TEDx Talk presentation here:
https://www.youtube-nocookie.com/embed/OdvErKtIVUE?rel=0&autoplay=0&showinfo=0&enablejsapi=0

