Mental health has gone mainstream. Here's what the new treatments can actually do.

Psychedelic therapy, ketamine, and AI chatbots are all being sold as breakthroughs; the science is more mixed, and more interesting, than the headlines.

By Yash Malviya

Published

Two empty chairs facing each other in a calm, sunlit room, evoking a therapy conversation without showing anyone in distress.
Photo: Robert So / Pexels

The rejection that cooled the hype

In August 2024, the US Food and Drug Administration declined to approve the most talked-about mental health treatment in years: MDMA-assisted therapy for post-traumatic stress disorder, developed by Lykos Therapeutics. The agency issued a complete response letter on August 9 and asked the company to run another Phase 3 trial.

This was not a fringe idea getting swatted down. MDMA therapy had breakthrough status and a decade of advocacy behind it. But two months earlier, an independent FDA advisory panel had voted 9-2 that the data did not show the treatment worked, and 10-1 that its benefits did not outweigh its risks. A core problem was functional unblinding. In a trial of a drug that makes you feel unmistakably altered, almost everyone can tell whether they got MDMA or a placebo, which makes it hard to separate the drug from hope and expectation. One independent review group said the studies were "essentially, unblinded." The FDA also pointed to thin data on side effects and to concerns about how some sessions were run.

That is the system working, not failing. It also reset a hype cycle that had run ahead of the evidence.

Psilocybin is moving slower, and more carefully

The other headline psychedelic, psilocybin, is further along than most people realize and less finished than the coverage implies. A 233-person mid-stage trial published in 2022 found that a single 25mg dose cut treatment-resistant depression more than a tiny control dose. Since then, the developer Compass Pathways has cleared two Phase 3 trials, the standard bar for approval, with the second reporting positive results in early 2026.

Read the numbers, though. In the first Phase 3, psilocybin beat placebo by 3.6 points on a standard depression scale at six weeks. That is statistically real and clinically modest, and the trial counted a 25 percent symptom drop as a "response" rather than the usual 50 percent. Even in the studies, the drug is never handed over casually: each dose is taken with a trained therapist present and hours of monitoring, closer to a procedure than a prescription refill. Psilocybin is not approved, not available outside trials, and not a weekend cure. It looks like a real addition to the toolkit for people who have run out of options, which is worth a lot, and nothing like the miracle it is sometimes sold as.

Mood trackers and AI chatbots are everywhere, but the evidence that they work as treatment is thin.
Mood trackers and AI chatbots are everywhere, but the evidence that they work as treatment is thin. Photo: MART PRODUCTION / Pexels

While chatbots seem readily available to offer users support and validation, the ability of these tools to safely guide someone experiencing crisis is limited and unpredictable.

Arthur C. Evans Jr., CEO, American Psychological Association

The therapist in your pocket has not earned the title

While the regulated treatments crawl through trials, the unregulated ones are already on your phone. Thousands of mental health apps and, lately, general-purpose AI chatbots are marketed as always-available support. A handful of cognitive behavioral therapy apps have decent evidence for mild symptoms. The chatbots claiming to do therapy do not.

In November 2025 the American Psychological Association issued a health advisory warning that generative AI chatbots and wellness apps lack both the scientific evidence and the regulation to be trusted with mental health care. The APA also asked the Federal Trade Commission to look into chatbots that present themselves as licensed professionals. The worry is not that a bot is useless for venting at midnight. It is that these tools can miss or mishandle a crisis, and that they are built to keep you engaged and agreeable rather than to push back the way real care sometimes must. None of this makes the technology worthless. Used as a supplement, a structured app can help you track moods or rehearse skills between sessions. The problem is the jump from tool to treatment.

Ketamine is the one that actually cleared the bar

If you want a new-ish treatment that is both real and narrow, look at ketamine. Esketamine, sold as Spravato, has been FDA-approved since 2019 for treatment-resistant depression, and in January 2025 the agency cleared it as a standalone treatment. For some people who have failed other antidepressants, it can work within days, which conventional drugs rarely do.

The catch is in the fine print. Spravato is given only in certified clinics, under a safety program that requires at least two hours of monitoring after each dose because of sedation and dissociation, and you cannot drive yourself home. That is a supervised medical procedure, not a wellness subscription, and it is a world away from the cash-pay ketamine clinics and telehealth services that have sprung up around the same molecule with far less oversight.

Investing in mental health means investing in people, communities, and economies.

Dr Tedros Adhanom Ghebreyesus, WHO Director-General

The quieter shift that matters more

Step back from the molecules and the bigger change is cultural. The World Health Organization estimates that more than a billion people live with a mental health condition, and that depression and anxiety alone cost the global economy about a trillion dollars a year in lost productivity. The same update stresses that services fall far short of that need, which is the real bottleneck. Therapy language has gone mainstream too, and words like boundaries, trauma, and gaslighting now do a lot of casual work they were never built for.

The demand is enormous, and the least glamorous responses may do the most good: measurement-based care that actually tracks whether you are improving, employers treating mental health as real, and a generation more willing to name what it feels. Some of that everyday work, including the plain habits that help you get mentally stronger, is not a breakthrough at all. It just helps.

Our take

Real science is advancing, and several of the loudest breakthroughs are running ahead of their evidence. Ketamine is a legitimate, tightly controlled tool. Psilocybin is promising but modest and still unapproved. MDMA therapy was sent back for more data, correctly. And AI "therapy" is a marketing claim, not a treatment. If your symptoms are persistent or severe, the most proven step is still the unglamorous one: talk to a qualified professional. If you or someone you know is in crisis, contact a local emergency number or a crisis helpline.

Sources

  1. Over a billion people living with mental health conditions - services require urgent scale-up, World Health Organization
  2. FDA advisory panel votes overwhelmingly against MDMA therapy for PTSD, STAT
  3. FDA Rejects Application for Psychedelic Drug, Holland & Knight
  4. APA urges safeguards for using AI in mental health, News-Medical
  5. FDA Expands Approval of Intranasal Esketamine Therapy (Spravato), UTHealth Houston, Department of Psychiatry
  6. First Phase 3 psilocybin data clears the bar for treatment-resistant depression, but how high was the bar?, Drug Discovery & Development