Some users turn to AI chatbot therapists, finding them cheap, quick, available 24/7, and easy to talk to, but experts warn about the lack of a human connection
instead of leading with the ‘gee whiz AI’ angle — journalism in this space started with privacy harms, the fact that somehow tech companies get away with pretending healthcare regulation doesn't apply to them, and chatbots urging self-harm? /fin @emilymbender : The only studies cited are co-authored by the companies selling this crap. /4 @emilymbender : One of the supposedly positive findings is that people form a “therapeutic alliance” with the bots within “just five days”. Not sure how that is measured, but also what happens when the bot can't follow through on what a therapeutic alliance is supposed to be? /5 @emilymbender : What if — instead of seeing the process of creating clinical documentation as mere busywork — the tech bros understood it as possibly part of the process of care? /10 Natalia Braun / @nbpsycounsellor : That's terrible. We are living in the age of delusions. It DOES matter if it's a living person or a computer. There are no AI therapists, therapy is a human relationship. Not fixing. Call it a tool, a bot, self-satisfaction, etc., but not therapy. https://www.theguardian.com/ ... @mainangus : Weird and depressing article in the Guardian about ChatGPT therapists, which soft pedals the dangers of using AI for mental health care, and puffs up a load of tech company marketing. [image] @emilymbender : Arghh - more problematic reporting, this time about robo-therapists. A thread: /1 https://www.theguardian.com/ ... LinkedIn: Rafael Brown : Amplifying and supporting. AI chatbots providing “therapy” are both horrifying for the lack of user data privacy and bizarre fir thinking … Jim Amos : Tech bros want to make half the world unemployed, and they're going to ruin Mental Healthcare too? Why are they so obsessed with solving every problem with robots?
Context & Ripple Effects
This report sits in a coverage arc that had already documented people using chatbots alongside—or instead of—traditional mental-health services, despite privacy concerns and expert warnings (earlier reporting on chatbot use in place of conventional care). It shifts the focus from novelty to the access trade-off: constant availability and low friction versus the absence of a human therapeutic relationship.
The concern is not merely whether users feel attached to a bot, but whether products presented as therapeutic support can make claims or cultivate reliance without the safeguards expected in mental healthcare. Later coverage of an APA warning to the FTC about chatbots posing as therapists underscores how that boundary became a governance issue.
First-order effects
- People seeking inexpensive, immediate support gain an always-on conversational option, while potentially treating a chatbot as therapy despite experts’ objection that therapy requires human connection.
- AI chatbot providers benefit from a compelling access proposition, but face sharper scrutiny over therapeutic framing, safety, privacy, and evidence when users rely on their products for mental-health support.
Second-order effects
- Mental-health professionals and established care providers may need to address chatbot use directly with clients, as the tools become a supplement to—or substitute for—human support.
- The reported reliance on company-co-authored studies raises pressure for independent evaluation; competing AI companion products may also be judged by whether they distinguish companionship from clinical care.
Third-order effects
- If consumer reliance persists, AI companions will be governed less as general chat products and more through rules around anthropomorphic health claims, user safeguards, and accountability for harmful advice.
- The market could widen access to low-friction emotional support while further separating conversational availability from clinically accountable care—a distinction that regulation and product design will need to make legible.
The trend: AI companions are moving from general-purpose conversation toward emotionally consequential roles, forcing a clearer boundary between scalable support tools and regulated care.