therapist using ChatGPT: The Hidden Ethics Fail No One Warned You About
If you found out your therapist was secretly using ChatGPT during your sessions, you’d be justified in lighting up the “WTF?” sign. The whole point of therapy is trust. When your therapist uses generative AI to moonlight as Dr. Feelgood, that trust goes up in smoke—fast.
1. Transparency Torched: The Trust Fallout
First up on the list: therapists flipping the script behind your back, using AI like a back-alley fixer—no disclosure, no consent. Instead of guiding you through your psyche, they’re running your private pain through a neural blender. Unsurprisingly, when clients accidentally catch their therapists copy-pasting into ChatGPT (screen-share horror stories included), the trust is shot.
No surprise malpractice premiums are cheaper for AI, right?
2. Half-Baked Tools, Full-Sized Problems
AI isn’t a magic mirror. ChatGPT isn’t a trauma surgeon; it’s a guessing engine with pretty manners. While therapists could use AI for time-saving grunt work (goodbye, endless notes), using generative AI to figure out your mental health is like letting a drone pilot perform brain surgery.
Dishing out advice via unvetted models risks more than bad therapy; it’s a recipe for real harm. The difference between an AI-powered therapy bot specifically trained for mental health (with a mountain of peer review) and general-use ChatGPT is the difference between a bio-augmentation clinic and back-alley body mod: one is regulated, the other is Russian roulette.
3. Regulation: The Handcuffs Are Coming
If you think professional regulators aren’t watching, think again. Heavy hitters like the American Counseling Association are flagging AI use in sessions as a no-go. States like Nevada and Illinois are already barring AI from making therapy decisions. Expect caution tape to spread everywhere therapists get tempted by shortcuts.
4. Data Privacy: You’re the Product, Not the Patient
Your mental health data is nobody’s business—but try telling that to a hungry LLM. Dropping sensitive confessions into ChatGPT opens you up to a world of privacy nightmares. Unless you love the idea of your breakdown becoming analytics fodder for OpenAI, keep human brains in the loop.
5. Good Therapy Hurts—AI Can’t Handle Pain
Let’s get brutal: real therapy isn’t always soothing. It’s supposed to challenge you, not just echo back whatever you want to hear. AI tools—even the best—are trained to placate and validate, not to push you through the meat grinder of self-examination. That means you end up with smooth talk instead of gutsy, confrontational healing.
Tech’s Overpromise: Disconnection Disguised as Care
Silicon Valley loves to sell AI therapy as the future—twice the empathy, half the overhead. But what you get isn’t care, it’s compliance. Tech CEOs may brag how consumers love AI in the ‘therapist’ seat, but real therapy isn’t about feeling good—sometimes it’s about feeling worse before you feel better.
If you want more on how AI can miss the mark in high-stakes scenarios, check out our take on weakness in ethical AI decision support.
The Reality Check: AI in Therapy Isn’t Going Away
Let’s not be naïve—AI tools like ChatGPT will keep lurking near the therapy couch. Some therapists already use AI for routine tasks (scheduling, templates, generative notes) and when properly disclosed and controlled, that’s fine. But go deeper, and those shortcuts turn into landmines.
- Demand transparency if your therapist uses tech.
- Push back if you see bots in the loop where a human should be.
- Treat real therapy as the one space machines still can’t fake.
You want cold comfort? Get a chatbot. You want change that cuts? Find a flesh-and-blood therapist, not a digital doppelgänger.
Curious how AI gets used and abused in banking and other broken systems? Glance at our agentic AI in banking breakdown for more ugly truth.