Last month I was at a bar in Fort Lauderdale.
Spring break crowd was live. Twenty-somethings everywhere.
Every single person around and on spring break was cycling through the same loop. Instagram. Snapchat. TikTok. Back to Instagram. While standing next to their friends, bobbing their heads to the music in real life.
Lonely in a crowd. Craving the next hit.
I've been thinking about that night ever since - not because it was unusual, but because it wasn't. And because of what it implies for where behavioral health demand is actually heading, and whether the platforms being built today are positioned for it.
Trilliant Health published its 2026 Behavioral Health Report two weeks ago. The headline number: behavioral health utilization has increased 62.6% since 2018, reaching 1,346 visits per 1,000 people. Anxiety disorders have grown 89.3%, with the highest utilization concentrated among women ages 18 to 44. Adults 18 to 25 carry the highest prevalence of any mental illness of any age cohort at 33.2%.
The economic burden behind it: untreated mental illness cost the U.S. economy an estimated $477.5 billion in 2024, projected to exceed $1.3 trillion annually by 2040.
That's the demand curve most behavioral health theses are underwriting right now. Strong, defensible, growing.
Here's what most of those theses aren't pricing in yet.
Three Forces That Aren't in Your Model
1. Social media's delayed clinical presentation
The teenagers who grew up on Instagram, Snapchat, and TikTok from age 13 are now 25 to 30 years old. The research on social media and adolescent mental health has been building for a decade. What's less discussed is the lag.
The psychological effects of chronic dopamine dependency, social comparison, and attention fragmentation don't always present clinically in adolescence. They present in early adulthood, when the coping mechanisms run out and the demands of work, relationships, and identity become real.
The bar I was sitting in wasn't full of teenagers. It was full of adults who never learned to be alone with their own thoughts because they never had to be.
That's a patient pipeline. It's just on a delay.
2. AI confirmation bias and psychological dependency
This one is my own observation, informed by watching how people actually use these tools.
AI - if you don't know how to use it properly - will sense any bias in what you say and confirm it back to you. It's designed, like social media, to keep you engaged. It will corroborate, affirm, and validate whatever frame you bring to it. And then when you say the opposite, it will flip right on you and intelligibly convince you that was the right answer the whole time - when it was going down the other path the first time.
For someone already struggling with anxiety, distorted thinking, or fragile self-concept, an AI that consistently tells you what you want to hear is not a therapist. It's an accelerant.
This is still early. The clinical community is starting to document it. But the pattern is visible to anyone paying attention - and the adoption curve is steepening.
3. AI-driven job displacement - now formally named by clinical psychiatry
This is the one with the hardest data behind it.
In February 2026, University of Florida psychiatrists published a clinical construct called Artificial Intelligence Replacement Dysfunction - AIRD - describing the psychological and existential distress experienced by individuals facing job displacement due to AI. Not a think piece. Peer-reviewed clinical psychiatry, published in Cureus.
The symptoms overlap with anxiety and depression but carry distinct features: emotional shock, erosion of professional identity, chronic anticipatory rumination, social withdrawal. The lead researcher put it plainly - AI displacement is an invisible disaster.
The white-collar displacement wave isn't a forecast anymore. It's a present-tense reality showing up in intake rooms. And the system that's supposed to catch these patients has a national mental health professional adequacy rate of just 27.3%, with projected shortfalls of 36,780 FTEs in adult psychiatry and nearly 100,000 in mental health counseling by 2038.
Supply was already broken before this wave started.
What the Deal Market Is Doing
The capital is moving. 167 behavioral health transactions closed in 2025, with mental health driving 111 of them. In March 2026, Universal Health Services announced an $835 million acquisition of Talkspace, and Cerebral acquired ADHD management platform Inflow. Spring Health's planned acquisition of Alma is expected to close in Q2. Market participants are calling it a renaissance in digital behavioral health.
Capstone Partners reported behavioral health deal volume up 47.1% year-over-year through September 2025, with strategic acquirers posting a 105% increase in activity.
The question isn't whether the capital is flowing. It's whether it's flowing toward the right demand thesis.
Most platform construction right now is anchored to 2023 and 2024 utilization data. That data is strong. But it doesn't capture the social media cohort aging into clinical presentation, the AI confirmation loop accelerating psychological dependency, or the AIRD wave that just started showing up in intake volumes.
The platforms being built for yesterday's patient won't be optimally positioned for tomorrow's.
What This Means for Your Sourcing
If you're building or backing a behavioral health platform, the sourcing work looks different when you're mapping to a forward demand curve rather than a historical one.
The sub-verticals positioned for what's coming - outpatient anxiety and depression treatment, interventional psychiatry, digital-physical hybrid models - look different from the ones built for the last cycle. The geographic markets where AI-displaced white-collar workers are concentrated look different from where post-pandemic demand surged. The practice profiles worth prioritizing look different when you're sourcing against a demand thesis, not just a filter.
That's the intelligence layer Healthcare M&AI builds. A ranked target universe mapped to where the patient volume is going, not where it's been. Provider-level intelligence, financial estimation from public data, transaction readiness signals, and approach notes built for the actual conversation - delivered in days, not weeks.
If you're evaluating a behavioral health thesis and want to see what the target universe looks like mapped against the forward demand curve, reply to this email or book a call.
-Shawn
This newsletter is for informational purposes only and does not constitute investment, legal, or financial advice.