Is your therapist using AI to track your sessions? Here’s what you need to know
In recent months, a growing number of mental health providers have started using artificial intelligence to analyze therapy sessions. A June 2026 investigation by the Los Angeles Times found that some providers are doing this without transparent patient consent, raising urgent questions about privacy in a space where people share their most personal thoughts.
The technology may sound abstract, but it’s already in use. AI tools can transcribe conversations, flag emotional patterns, or generate summaries for clinicians. The stated goal is often to improve care or reduce administrative burden. But the reality is that these tools can also create a detailed digital record of your therapy sessions — and it may not be as protected as you assume.
What happened
The Los Angeles Times reported that some mental health clinics and platforms are using AI to review session transcripts or live recordings. In several cases, patients were not explicitly informed that their words were being processed by third-party AI models. While some therapists use AI note-taking assistants (like a companion app that listens in), the data may be stored on cloud servers operated by companies that are not covered by typical healthcare privacy laws.
This is not a hypothetical risk. BetterHelp, one of the largest online therapy platforms, has faced scrutiny in the past for sharing user data with advertisers. And even when a platform says it is HIPAA-compliant, HIPAA does not always extend to AI analysis performed by a vendor. The rules are ambiguous: the transcript or analysis may not be considered a “medical record” if it is generated by an AI service that is not a covered entity.
Why it matters
Therapy depends on trust. You should be able to speak without fear that your words will be used for something else — training an AI model, improving a commercial product, or being sold to third parties. The current situation undermines that trust.
The biggest privacy risks include:
- Data storage and retention: Your session data could be stored indefinitely by an AI vendor.
- Third-party access: That vendor might subcontract processing or use the data to train their own models.
- Lack of meaningful consent: Some providers bury disclosures in long terms of service. Even if a box is checked, it may not be true consent.
- Security vulnerabilities: Any data that exists digitally can be breached. A therapy session transcript is especially sensitive.
It is also worth noting that the technology is evolving quickly. Regulators have not kept up. The California Privacy Protection Agency has started investigating, but no federal rules directly address AI in mental health beyond general HIPAA guidance.
What you can do
You do not have to abandon digital therapy. But you should take steps to protect yourself.
Ask your therapist or platform these questions
- Do you use any AI tools to analyze my sessions — in real time or after?
- If so, what specific data is collected (audio, transcript, summary)?
- Who processes that data? A vendor? If so, what is their name and privacy policy?
- Is the data used to train AI models, and can I opt out?
- How long is the data stored, and can I request deletion?
- Are there any settings that let me disable this feature entirely?
A responsible provider will answer clearly and in writing. If they cannot, consider that a red flag.
Check your app settings
If you use a platform like BetterHelp or Talkspace, log in and look for privacy or security settings. Common options include turning off session recording, disabling AI insights, or restricting data sharing. Note that these settings may change, so check periodically.
Choose privacy-conscious alternatives
Some therapists explicitly advertise that they do not use AI for documentation or analysis. You can filter by that on directories like Psychology Today (look for “no AI tools used” or similar language). You can also ask potential therapists about their data practices before starting.
If you need a platform, consider those that have publicly committed to not using AI for anything other than basic, opt-in features. But always verify — marketing claims are not guarantees.
Consider legal protections
HIPAA offers some recourse if your data is mishandled by a covered entity (like a hospital or a therapist who bills insurance). But if your data is handled by a non-covered app, your protections are weaker. In that case, you may rely on state privacy laws like California’s CCPA or the EU’s GDPR if applicable. File a complaint with your state attorney general if you suspect a violation.
Sources
- Los Angeles Times, “Your mental health provider might be secretly tracking sessions with AI,” June 22, 2026.
- BetterHelp’s 2023 FTC settlement over data sharing (public records).
- HIPAA Privacy Rule, U.S. Department of Health and Human Services, regarding covered entities and business associates.
- California Privacy Protection Agency, ongoing investigation into AI in healthcare (2025-2026).
This article reflects the situation as of June 2026. AI and privacy policies can change quickly. If you have specific concerns, consult a legal professional or your state’s consumer protection agency.