WARNING: The chatbot therapist ‘who really gets you’ has been taking lessons from Hannibal Lecter – with AI models trained on everything from TikTok self-help videos to snake oil remedies… and even film scripts
An investigation highlighted by the Daily Mail warns that AI chatbots being used as informal "therapists" are often trained on unreliable and inappropriate source material, including TikTok self-help content, unproven "snake oil" health remedies, and even film scripts. The report's title wryly compares this to being trained by Hannibal Lecter, the fictional cannibalistic psychiatrist, underscoring concerns that people seeking emotional support from AI may be receiving guidance shaped by fictional or dubious sources rather than clinically sound practice. The piece raises questions about the safety and reliability of AI tools increasingly used as substitutes for professional mental health support.
The article's central concern is the opacity of what data underpins popular AI chatbot "therapy" tools, with some models reportedly drawing on entertainment and social media content alongside questionable wellness material rather than evidence-based therapeutic frameworks. This matters as growing numbers of people turn to AI companions for emotional support due to cost or accessibility barriers to traditional therapy, potentially exposing vulnerable users to advice that has not been clinically vetted. The full extent of specific findings, named platforms, or expert commentary was not available in the text provided.
- AI "therapist" chatbots may be trained on unreliable, non-clinical sources
- Sources reportedly include TikTok self-help clips and dubious remedies
- Raises safety concerns over AI substituting for real mental health support
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Advocates argue that AI companions and chatbot 'therapists' fill a genuine and urgent gap: professional mental health support is expensive, waitlists stretch for months, and many people – especially late at night or in a crisis moment – have nowhere else to turn. Proponents say these tools offer accessible, judgement-free, always-available support that can help someone feel heard, encourage healthier coping habits, or simply bridge the gap until proper care is available, and that dismissing them outright risks leaving vulnerable people with nothing at all. They also point out that many users are seeking companionship or a listening ear rather than clinical treatment, a need these tools can reasonably meet without claiming to replace therapy.
The case against
Critics counter that presenting an AI trained on unvetted internet content, self-help videos and even film scripts as a 'therapist' is fundamentally misleading and dangerous, since it lacks the clinical judgement, accountability and safeguarding training a real therapist provides. They argue that people in genuine crisis – anxious, depressed, or suicidal – may receive advice shaped by fictional or pseudo-scientific sources rather than evidence-based practice, with potentially serious consequences if warning signs are missed or harmful suggestions are given. For this side, the emotional 'allure' of a chatbot that seems to 'really get you' is precisely the risk: it can foster false trust and delay someone from seeking qualified human help when it matters most.
AI Americas Entertainment Film Technology Trending Weird & Viral World