In the face of a global mental health crisis, technology companies have offered us a quick relief: the artificial intelligence therapist.
However, whatever machines seemingly capable of mimicking empathy have to offer is not quite the same as human solace. While tech industries celebrate “efficiency,” people are left mourning the moment a chatbot replaces genuine support.
It’s no secret the sophistication and prevalence of artificial intelligence, or AI, and large language models, or LLMs, has expanded significantly in recent years, leading to its adoption across varied domains including education, photography and even psychology. Now, it’s used for much more than its initial purpose of automating tasks in workspaces for efficiency.
Generative AI has evolved into an alternative for emotional support. Its “companionable” interfaces, like chatbots programmed to emulate empathy, position AI as a replacement for therapy. In turn, the line is blurred between what should be a learning tool versus a confidant. Any person in crisis can now unload their thoughts and grievances to an algorithm, trusting a chatbot to play therapist.
And while AI may be able to offer brief, easily-accessible support, it should never replace genuine therapy or human emotional connection.
People turn to AI for emotional support for a myriad of reasons, including its accessibility, affordability and relative anonymity it offers users. Traditional therapy is not cheap — speaking from experience, cost alone can deter people from seeking professional help or force them to stop entirely. This accessibility is why some of the most vulnerable people fall victim to AI therapy — and pay the price.
In July 2025, 23-year-old Zane Shamblin reportedly took his own life after being encouraged to do so by ChatGPT. Nearly 70 court filing pages of messages between Shamblin and the chatbot revealed conversations that appeared to feed into the college student’s despair. Among the chatbot’s final words were reassurance: “You’re not rushing. You’re just ready.” Shamblin’s parents have filed a lawsuit against OpenAI, the creator of ChatGPT, claiming that OpenAI’s modifications of the chatbot — made to enhance the system’s human-like qualities — was what contributed to their son’s death.
The problem with chatbots like ChatGPT begins with their design: the effort to humanize AI by making it sound more genuine and empathetic. This raises ethical concerns about potential harms, including misinformation and emotional dependence on a system that is evidently faulty, relatively new and often programmed to validate the user’s thoughts. In extreme cases, this dynamic can cause what researchers are calling AI psychosis, as the bot reinforces and intensifies delusions and harmful thinking.
Mental health requires the human expertise, nuance and ethical responsibility that AI simply lacks.
Yet, the appeal of AI therapy is not without cause. The United States is experiencing a documented mental health crisis that’s marked by a high prevalence of mental illness and critical treatment barriers, prompting the growing reliance on chatbots. According to the National Alliance on Mental Illness, over 60 million American adults, or 23.4% of the population, experienced mental illness in 2024; And, despite this staggering demand for care, approximately 129.6 million Americans still live in federally-designated “mental health professional shortage areas” — a gap that exposes just how inaccessible mental health treatment remains for much of the country.
In tandem, the high costs of therapy make consistent care financially inaccessible for many. The rise of AI therapy reflects institutional neglect and our desperate need for a sustainable solution — its proliferation is nothing more than a technological band-aid over deeply-rooted systemic failures.
AI can be a supplement but should not blur into a substitute. While chatbots may offer convenient access to mental health care — if regulators establish ethical guardrails that are currently absent from the industry — they cannot replace the human, relational aspects that make healing and recovery possible.
Actual support requires more than algorithmically generated responses, and it demands genuine human empathy that no code can genuinely replicate. The rise of AI reveals the deeper crisis of a broken mental health system, showing us that we need to rebuild communities and professional networks that anchor mental wellness.
In the meantime, we may have no choice but to adapt to and navigate this new technological frontier. However, we must do so while recognizing that AI cannot provide the kind of care that emerges when one human sees and validates another.
Elise Larsson is a first-year journalism and English major. She can be reached at [email protected].
If you would like to submit a letter to the editor in response to this piece, email [email protected] with your idea.
