Audio Version: Audio player to be added by SCN before publication.

A teenager types something into a mental health app at 11 at night, hopes it helps, closes the phone, and goes to sleep. Nobody else in the house knows where those words went.

This is not a therapist’s office with a locked filing cabinet. Most mental health apps are not bound by the rules that protect a therapy session, and a study published this year found that every one of the 25 popular apps it tested carried at least one tracker its own privacy policy never named.

Quick Review

  • Every one of the 25 apps carried at least one tracker its privacy policy did not name.
  • Nineteen of 25 asked for camera access, microphone access, or both.
  • Only seven let a user delete data inside the app.
  • Twelve said in their policies that data goes to third-party AI providers. Seven of those would not name a single company.
  • A therapist-backed service works under different rules than a self-help app. That difference is the first thing to check.

What Mental Health Apps Ask Before Your Teen Even Signs Up

The study is “What’s on Your Mind? Exploring Privacy of Mental Health Apps”, by Chloe Georgiou, Hans Lu, Emiliano De Cristofaro, and Gene Tsudik of the University of California, Irvine and the University of California, Riverside. Submitted in May 2026 and revised in August, it examined 25 popular Android mental health and life-coaching apps, from Headspace and Replika to Talkie, 7 Cups, Wysa, Youper, and BetterUp. Several have more than 10 million downloads.

Of the 20 apps researchers could open without an organizational access code, 16 asked for an email address and nine asked mental health screening questions before an account existed. That means a teen may share how they feel before reading the privacy policy. Three apps required gender, age, race, and mental health information for full use.

The Tracker in Every Single App

Researchers counted 132 tracker instances from 37 distinct trackers across the 25 apps. Only 18 tracker instances, about 14 percent, appeared in a privacy policy. At runtime, every app the team could run contacted at least one third-party tracking domain the policy did not mention. The Facebook SDK appeared in 10 apps, and none of those apps’ policies disclosed it.

Why does that matter? An advertising network does not need to read a teen’s messages to learn they are using a therapy app. The advertising ID on a phone is shared across every app on it, so the same tracker inside a mood app and inside a shopping app can connect the two. Research cited in the paper found data brokers selling lists of people sorted by mental health condition.

None of that proves a company read your teen’s entries. It does mean the app shared more about why they opened it than the policy admits.

The Camera and Microphone Question

Nineteen apps declared camera access, microphone access, or both. Six requested camera or microphone access without disclosing it in their policies: Aura, Hapday, Rocky, Rosebud, Sensa, and Yana.

Wysa, Youper, Happify, and Aura told users during onboarding that their answers were private or confidential, while embedding trackers their policies did not name. The study’s authors cautioned that their findings are not evidence of deliberate misconduct. Gaps may come from inherited software kits or boilerplate policy text, and the researchers did not accuse any company.

Where the Words Go, and Who Can Read Them

Twelve apps said in their policies that data goes to third-party AI providers. Only five named a company; the names included OpenAI, Anthropic, Groq, and Google Gemini. Seven used vague phrases such as “AI services” or “large language models.”

Some policies allow user content to train models. Noah, Pi, and Rosebud mention training on user content. Rosebud’s policy says the resulting datasets and trained models may be sold on to commercial AI laboratories.

In every chatbot app the researchers could operate and decrypt, text entered in the chat left the phone for the company’s own servers. That is how these apps work, but it is worth saying plainly to a teen before they share something personal. Five apps whose main feature is a chatbot never mention AI processing in their policy at all.

Every app in a 25-app study carried at least one tracker its own privacy policy never named.

Know a parent who needs this? Send it to them.

What a Therapist-Backed Service Promises Instead

A licensed clinical service is different from a self-help app. New York City’s Teenspace program, a free teletherapy service run with Talkspace under a $26 million, three-year contract, had roughly 52,000 teens signed up as of April 2026. Its future is in limbo while the health department conducts an independent evaluation, Healthbeat reported on November 11, 2026.

Privacy advocates had raised concerns about data being shared with third parties. After those concerns, Talkspace agreed in December 2024 to remove tracking pixels from the Teenspace website that were not needed for it to work. The city health department says data is protected and used only to provide the service; the contract requires data to be destroyed after 30 days if a teen never starts therapy.

The Teenspace privacy policy states that a teen’s identifying information and medical information are not sold and are not used for marketing, advertising, or commercial purposes, and that medical information is protected health information under HIPAA. It also says Teenspace pages carry only the cookies and pixels needed to run and secure the site, with no social media or advertising trackers. The lesson is not that one company is good and another bad. It is that a licensed clinical service operates under rules a mood journal does not.

That difference has teeth. In 2023 the Federal Trade Commission ordered the online therapy company BetterHelp to pay $7.8 million and banned it from sharing health information for advertising, after the agency said it revealed sensitive mental health data to Facebook, Snapchat, Criteo, and Pinterest while promising users confidentiality. BetterHelp is not one of the 25 apps in this study. It is the clearest recent reminder of what a privacy promise is supposed to mean.

A parent and a teenager sitting side by side at a dining table, both looking at a laptop, with a fruit bowl and two mugs on the table and afternoon light coming through the window.

Five Questions To Ask Before Your Teen Downloads

Ask these questions together. Keep the conversation practical and calm.

  1. “Is a licensed therapist or clinic behind this app, or is it self-help with a chatbot?”
  2. “Does the privacy policy name the analytics companies it uses, or just say ‘third parties’?”
  3. “Can you delete the account and data inside the app?” Only seven of the 25 apps offered in-app deletion; 19 required an email request, and CBThelp and Clarity stated that they had no deletion mechanism.
  4. “Does the policy say whether what you type can train an AI model?”
  5. “Does the app ask for camera or microphone access, and if so, why?”

Apps outside HIPAA may still fall under the FTC’s Health Breach Notification Rule. Since its 2024 amendments, the rule covers health apps not covered by HIPAA, treats unauthorized disclosure of health data as a breach, and carries penalties of up to $53,088 per violation. Earlier FTC health app cases against GoodRx and the ovulation app Premom ended in $1.5 million and $100,000 penalties. Those were cases involving those companies, not mental health apps.

Separately, a proposed class action filed November 5, 2026, in the Northern District of California, Mosquera v. Opal OS Corporation, alleges a screen-time and habit app sent onboarding answers and app activity to analytics company Amplitude without consent. Opal has not responded in court, and no class has been certified. The complaint and its allegations are not proven facts.

For a broader family safety plan, explore SCN’s guides to parental controls and their limits and AI versus traditional parental controls.

What To Tell Your Teen About What They Type

Try: “The app may be helpful, and the privacy question is separate. Let’s check what it does with your information together.”

A paper journal beside the bed is a real option. Your teen never has to tell you the details of what they write to deserve your support. If they are in crisis, they can call or text 988, the Suicide and Crisis Lifeline.

A vertical cream pin with deep plum type and simple plum line drawings of a phone with a chat bubble, and an open padlock with a question mark beside it.

Frequently Asked Questions

Are Mental Health Apps Covered By HIPAA?

Not necessarily. HIPAA applies to covered health care providers and their business associates, not automatically to every self-help or wellness app. Check whether a licensed provider is involved and read the app’s privacy policy.

Can My Teen Delete Their Data From A Mental Health App?

It depends on the app. In the study, only seven offered in-app deletion; some required an email request, and CBThelp and Clarity stated no deletion mechanism. Check the policy and the app’s account settings before signing up.

Is It Safe To Let My Teen Use An AI Chatbot For Anxiety?

The study found that some apps send chat text to company servers and that policies vary in how they describe AI processing. Review what the service collects, who receives data, and whether content may train models, then decide together what your teen is comfortable sharing.

What will you check first when your teen considers a mental health or wellness app?

For practical guidance on children’s digital well-being, join the Secure Children’s Network newsletter.

If this helped, share it with another parent and follow Secure Children's Network on Facebook, X and LinkedIn for safety alerts.

Author: Janice Robinson-Celeste


Discover more from

Subscribe to get the latest posts sent to your email.

Leave a comment

Discover more from

Subscribe now to keep reading and get access to the full archive.

Continue reading

Secure Children's Network logo
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.