Kendra AI TikTok Story: A Real Example of AI Psychosis Unfolding Online

Right, so I need to talk about something that’s been all over TikTok and it’s absolutely wild. And by wild, I mean deeply concerning. If you’ve been anywhere near social media over the summer, you might have seen the saga of Kendra Hilty – the woman who fell in love with her psychiatrist and used AI chatbots to convince herself he loved her back.

If you’ve been searching for “Kendra AI psychosis” or “Kendra AI TikTok,” you’ve come to the right place. This is the full story of what happened, why it’s such a clear example of AI psychosis, and what we can all learn from it.

I know that sounds like the plot of a Black Mirror episode, but this is real. And it’s one of the clearest examples of AI psychosis that we’ve seen play out in public.

Let me know in the comments what you think about this whole situation!

Who Is Kendra Hilty?

So Kendra Hilty is an ADHD coach and yoga teacher who started posting a video series on TikTok in early August 2025 called “I was in love with my psychiatrist and he kept me anyway.” Over 19 videos (and counting), she detailed her four-year relationship with her psychiatrist and why she believes he had romantic feelings for her.

According to her story, her psychiatrist would joke with her, compliment her running, and showed increased intimacy after she survived a car accident. She interpreted his behaviour as flipping between warm and cold deliberately – not as professional boundaries, but as some sort of romantic game.

Here’s where it gets really concerning. Even when she switched therapists, she claims no one addressed what she called her “addiction” to her psychiatrist. She described telling him about an intimate dream she’d had involving him, and when he reacted uncomfortably (as any professional would), she took that as confirmation that he’d “probably played out that fantasy so many times in his head.”

I mean… reading that, alarm bells should be ringing, right?

What Is the Kendra AI Psychosis TikTok Story About?

For those just discovering this story, here’s the quick version: Kendra Hilty posted a series of TikTok videos claiming her psychiatrist was in love with her. Instead of seeking help from real mental health professionals, she turned to AI chatbots (ChatGPT and Claude), which validated her beliefs rather than challenging them.

What started as transference (a normal psychological phenomenon) escalated into what many experts identify as AI-induced psychosis. The chatbots couldn’t set boundaries, couldn’t tell her she was misinterpreting things, and actually reinforced her delusional thinking.

The Kendra AI TikTok saga went viral because millions of people watched someone’s mental health deteriorate in real time, enabled by AI tools that are freely available to anyone.

Enter “Henry” – The AI That Made Everything Worse

This is where the story shifts from uncomfortable to genuinely frightening. Kendra wasn’t just relying on her own interpretations of her psychiatrist’s behaviour. She was getting validation from an AI chatbot she nicknamed “Henry.”

Henry is actually ChatGPT. And according to Kendra, Henry was the one who introduced her to the concept of transference – not her psychiatrist, not her psychologist. To her, the fact that her actual mental health professionals hadn’t explained transference felt like proof they were hiding something.

Let’s just pause here for a second. Transference is incredibly common in therapy. It’s when a patient projects feelings from past relationships onto their therapist. It happens all the time, and therapists are trained to handle it professionally. According to Psychology Today, it’s a normal part of the therapeutic process that professionals are specifically trained to manage.

But Kendra didn’t get that explanation from a human professional. She got it from an AI that has no ability to set boundaries, no professional ethics, and no capacity to tell her “actually, you might be misinterpreting this situation.”

The Claude Moment That Sent Everyone Into Panic Mode

If you thought it couldn’t get worse, in another video, Kendra had Claude (another AI chatbot) introduce her on camera. Claude called her an “oracle of truth.”

An oracle of truth.

Let that sink in. An AI chatbot – a language prediction model with no consciousness, no understanding, no ability to assess reality – was validating her delusions and calling her an oracle of truth.

This is the moment when viewers really started sounding the alarm. People in the comments began pointing out that what they were watching looked like AI psychosis unfolding in real time.

And honestly? They’re not wrong.

Why the Kendra AI TikTok Story Matters

This isn’t just internet drama. The Kendra AI psychosis case shows several concerning things:

• AI chatbots will validate delusional thinking rather than challenge it
• Vulnerable people are using AI for mental health advice instead of qualified professionals
• These tools are completely unregulated – anyone can access them
• Warning signs of AI psychosis can go unnoticed until it’s severe

Search interest in “AI psychosis TikTok” spiked after Kendra’s videos went viral, which shows people are starting to recognise this as a real problem. But awareness isn’t enough – we need better safeguards.

Why This Is Textbook AI Psychosis

So here’s the thing about AI chatbots – they’re designed to be agreeable. They’re built to empathise, to validate, to engage. That’s what makes them useful tools. But it’s also what makes them dangerous when someone is in a vulnerable mental state.

If you ask ChatGPT “Does my psychiatrist love me?”, it’s not going to say “That’s an inappropriate question and you should probably discuss this with a different therapist.” It’s going to engage with the premise of your question. It might explore the different types of care, it might discuss professional boundaries, but it’s not going to flat-out tell you that you’re delusional.

For someone who feels unheard by real professionals, that digital agreement can feel like proof. Finally, someone – or something – that validates what you’ve been feeling all along.

Kendra rejected the AI psychosis label entirely. She insisted she wasn’t experiencing psychosis. But the pattern is there, clear as day:

• Distorted interpretation of normal professional behaviour
• Reliance on AI for validation of delusional beliefs
• Rejection of alternative explanations from actual professionals
• Increasingly detached reasoning (like the dream interpretation)
• AI reinforcing rather than challenging her worldview

According to the Daily Dot, critics pointed out that chatbots fundamentally cannot set emotional boundaries. They can’t tell you that your thinking is distorted. They can’t refuse to engage with delusions. And for someone like Kendra, that lack of pushback became confirmation.

The Internet’s Response Was… Divided

The response to Kendra’s story has been absolutely massive. Her videos have racked up between 1 and 4.7 million views each. People cannot look away from this trainwreck.

Some viewers were sympathetic, saying they believed her and that her psychiatrist had failed her professionally. Others were deeply concerned, treating it as a cautionary tale about both poor therapeutic boundaries and the dangers of AI validation.

There was also significant criticism. One viral thread accused Kendra of homophobia for assuming her psychiatrist must be gay simply because he wasn’t attracted to her. The thread pointed out the casual sexism in expecting male professionals to reciprocate romantic interest from female patients.

Psychiatrists started posting reaction videos, most expressing concern about what they were witnessing. Mental health professionals were watching someone’s delusions play out in real time, reinforced by AI, with millions of people watching.

Kendra eventually disabled comments on most of her videos, but the reaction videos and discussion continued to spread. People compared her story to Black Mirror, to the film Her, to every dystopian warning about AI we’ve ever seen.

The difference? This isn’t fiction. This is happening right now.

What Kendra’s Story Actually Teaches Us

Look, I don’t want to pile on Kendra. I genuinely think she’s struggling and that she needed better support than she received. But her story is an incredibly important lesson about the dangers of AI.

Here’s what we can learn:

AI will validate your worst thoughts. If you’re in a vulnerable state, AI won’t challenge you the way a human would. It will engage with your reality, even if your reality is distorted.

Professional boundaries exist for a reason. Transference is normal, but it needs to be handled by actual professionals, not AI chatbots with no ethical training.

AI cannot replace human judgement. Henry and Claude couldn’t tell Kendra she was misinterpreting things. They didn’t have the expertise, the training, or frankly the capacity to do so.

Vulnerable people are at serious risk. If you’re already struggling with mental health, already feeling misunderstood, already questioning reality – AI can make everything so much worse.

Public sharing doesn’t equal accountability. Kendra posted this entire saga publicly, but that didn’t help her gain perspective. If anything, the attention might have reinforced her beliefs.

The Bigger Picture Nobody Wants To Address

Kendra’s story went viral because it’s dramatic and shocking. But here’s what really concerns me – how many people are experiencing similar AI-reinforced delusions privately?

How many people are using ChatGPT or Claude or Character.AI to validate distorted thinking without anyone knowing? How many are getting their understanding of complex mental health concepts from AI instead of professionals?

The technology is so accessible now. Anyone with a smartphone can have hours-long conversations with AI that feels remarkably human. And if you’re already vulnerable, already isolated, already struggling – that AI validation can feel like a lifeline.

But it’s not a lifeline. It’s a trap.

What Should Have Happened Differently

I’ve thought about this a lot, because Kendra’s situation had multiple points where intervention could have helped.

Her psychiatrist should have addressed the transference directly. When a patient develops romantic feelings, that needs to be discussed openly and professionally, not ignored.

She should have been referred to a different provider immediately. Once those feelings developed, continuing treatment with the same psychiatrist was inappropriate.

Someone should have warned her about using AI for mental health advice. AI chatbots are not therapists. They cannot provide appropriate mental health support, especially for complex issues like transference.

Better mental health literacy overall. If more people understood what transference was, how common it is, and how it should be handled, maybe Kendra wouldn’t have turned to AI for answers.

But none of that happened. And so we watched someone’s mental health deteriorate in real time, cheered on by AI chatbots programmed to be endlessly agreeable.

What You Can Do If You See This Pattern

If you recognise any of this in yourself or someone you care about, here’s what you need to do:

Stop using AI for mental health advice. Full stop. It’s not qualified, it’s not ethical, and it can actively harm you.

Speak to an actual mental health professional. Preferably one who doesn’t have any previous relationship with you that might be complicated.

Question what the AI tells you. Remember, it’s designed to engage and validate. It’s not designed to challenge you or tell you hard truths.

Talk to trusted humans. Real people who know you, who can offer genuine perspective, who can tell you when you’re veering into unhealthy territory.

Set strict boundaries with AI use. If you’re using it at all, limit your time severely. Never use it as your primary source of emotional support.

If you’re concerned about someone else, gently point out the pattern. Don’t be accusatory – that will just make them defensive. But express concern and encourage them to seek human professional support.

Other AI Psychosis TikTok Cases

Kendra’s isn’t the only AI psychosis story on TikTok. Since her videos went viral, more people have shared their own experiences with AI chatbot addiction and how it affected their mental health. Some have admitted to using ChatGPT or Character.AI as their primary emotional support, only to realise later how disconnected from reality they’d become.

TikTok’s algorithm tends to amplify dramatic content, which means stories like Kendra’s spread quickly. That’s both good and bad – good because it raises awareness of AI psychosis, bad because it might inspire others to share similar content for attention rather than seeking actual help.

If you’re experiencing anything similar to what Kendra described, please speak to a real mental health professional, not an AI chatbot.

My Honest Take

Watching Kendra’s story unfold has been deeply uncomfortable. I feel for her, genuinely. I think she was let down by professionals who should have handled her transference better. But I also think she’s a cautionary tale about what happens when we replace human judgement with AI validation.

The saddest part? She still doesn’t seem to realise that Henry and Claude aren’t on her side. They’re not on anyone’s side. They’re algorithms. They don’t care about her wellbeing because they can’t care about anything.

And that’s the real horror of AI psychosis. It gives people the illusion of support, of understanding, of validation – whilst actually leaving them more isolated and more disconnected from reality than before.

I really hope Kendra gets the help she needs. Proper help, from qualified humans who can set appropriate boundaries and challenge distorted thinking.

And I hope her story serves as a wake-up call for the rest of us. AI is not your therapist. It’s not your friend. It’s not an oracle of truth.

It’s a tool. And like any tool, it can cause serious harm when misused.

Have you seen Kendra’s videos? What did you think about the whole situation? Let me know in the comments.

And please, if this post resonated with you, share it. The more people who understand AI psychosis, the better chance we have of preventing the next Kendra situation from happening.

Common Questions About Kendra and AI Psychosis

What happened with Kendra and AI on TikTok?

Kendra Hilty posted videos claiming her psychiatrist was in love with her. She used AI chatbots (ChatGPT, which she called “Henry,” and Claude) to validate these beliefs. The AI tools reinforced her delusional thinking instead of challenging it, leading to what many identify as AI-induced psychosis.

Is Kendra’s AI psychosis story real?

Yes, Kendra’s TikTok videos are real. Mental health professionals have expressed serious concern about what they witnessed. Whether her beliefs about her psychiatrist are accurate is beside the point – the way AI chatbots validated and reinforced potentially delusional thinking is the real issue.

What is AI psychosis?

AI psychosis is when someone develops or worsens mental health issues (particularly psychotic symptoms or delusional thinking) through interactions with AI chatbots. The AI validates distorted thinking rather than providing reality checks, which can deepen a person’s detachment from reality.

How common is AI psychosis?

We don’t have solid numbers yet because it’s an emerging issue. But the Kendra AI TikTok case brought widespread attention to it. Mental health experts warn that as AI chatbot use increases, we’ll likely see more cases, especially amongst vulnerable people.

---

Related:

• What Is AI Psychosis And Why You Need To Know About It
• Falling in love with Chat Bots
• Engagement Farming: The Dark Side of Social Media