Millions of people are managing their mental health through apps built by people who have never spent a day in a clinical setting. That matters. Because when the algorithm gets it wrong, the cost isn’t a buggy playlist — it’s someone’s psychological wellbeing on the line.
The American Psychological Association has been paying closer attention to what’s actually inside these apps, and their findings — detailed in a recent breakdown of red flags psychologists watch for in mental health apps — should make every casual Calm subscriber pause before their next breathing exercise.
The App Store Is Not a Therapist’s Office
There are over 10,000 mental health apps available across major platforms right now. Ten thousand. And the barrier to publishing one is essentially a developer account and a decent logo. No clinical review board. No peer-reviewed research requirement. Just a smooth onboarding flow and a subscription prompt.
Psychologists have started building out a real framework for evaluating these tools, and what they’re finding isn’t reassuring. The red flags aren’t always obvious. They don’t show up as glitchy UI or bad star ratings. They show up as apps that overpromise outcomes, collect sensitive data without transparent policies, and push users toward engagement rather than recovery.
That last one deserves more attention than it gets. Engagement and healing are not the same goal. An app optimized to keep you opening it daily isn’t necessarily an app that wants you to need it less. Addiction mechanics dressed up in wellness language are still addiction mechanics.
What the Red Flags Actually Look Like
Vague or Absent Clinical Credentials
If an app claims to be “clinically backed” but can’t point to published, peer-reviewed research — not a white paper they funded themselves — that’s a problem. Marketing language is cheap. Real clinical validation is hard, slow, and expensive. Apps that skip it and still slap the word “clinical” on their homepage are lying to users who are often already vulnerable.
Data Practices That Should Make You Nervous
Mental health apps collect some of the most intimate data that exists. Mood logs. Sleep patterns. Thought records. Crisis check-ins. And too many of them have privacy policies that are either incomprehensible or quietly permissive about third-party sharing. If you’re already worried about what AI tools know about you and how they use it, your mental health app deserves at least the same level of scrutiny — probably more.
No Crisis Protocols Worth Mentioning
An app that addresses anxiety or depression but has no clear, functional path to emergency support is not a mental health app. It’s a mood journal with better design. When someone in genuine distress opens that app at 2am, a breathing exercise and a motivational quote are not sufficient. Psychologists flagging this aren’t being overcautious — they’re identifying a gap that has real consequences.
The Hot Take
Most mental health apps should not exist. Not because the category is bad — the idea of accessible, affordable mental health support is genuinely worth pursuing — but because the execution is, across the board, driven by venture capital timelines rather than clinical responsibility. The best-funded apps in this space were built to scale, not to heal. They were built to hit user growth targets, not remission rates. Until the incentive structure changes, until there’s real regulatory pressure or clinical gatekeeping baked into the app store approval process, we are essentially letting startups experiment on mentally ill people in real time. That’s not hyperbole. That’s the business model.
So What Do You Actually Do?
Psychologists recommend treating these tools the same way you’d treat a supplement rather than a prescription. Potentially useful, worth researching, but not a replacement for actual care. Look for apps that cite specific studies you can look up yourself. Look for transparent data policies written in plain language. Look for crisis resources that are more than a hotline number buried in the settings menu.
And be honest about what an app can and can’t do. Apps are good at habit support, psychoeducation, and low-stakes skill building. They are not good at diagnosis, trauma processing, or replacing the kind of relational attunement that actual therapy provides. Knowing the difference protects you.
It’s a little like what happens when technology tries to make sense of genuinely complex systems — sometimes the output looks impressive until you look closer. The same way AI can detect thousands of microquakes along fault lines but still can’t predict the big one, these apps can surface patterns in your mood data without ever really understanding what’s underneath them.
The mental health app market isn’t going away. The demand is real, the access gap is real, and for a lot of people an imperfect app is better than nothing at all. But “better than nothing” is a low bar. People dealing with depression and anxiety and trauma deserve tools held to a higher standard than the one currently in place — and right now, it’s psychologists doing the work that regulators and app stores have failed to do.