Trends in Behavioral Health Tech to Watch for in 2025
By Loren Larsen
The convergence of technology and behavioral health has seen rapid acceleration in recent years. Between the growing appeal of digital therapeutic apps, the evolution of clinical note-taking tools, and the growing ability of AI tools to meet patients’ and providers’ needs, many of these trends are already in motion.
Here’s what to watch for on these frontiers and more in 2025:
- Continued adoption of clinical note-taking tools. Technology is often met with distrust among practitioners. In the case of automated note-taking tools, clinicians are eager to begin using them, even fighting to be first in line within organizations that implement staggered rollouts. The number of HIPAA-compliant note-taking tools is growing every day. Already, they have made a huge impact on the organizations and practitioners that have adopted them—particularly in the mental/behavioral health space, where conversations are essential to the treatment process. These tools are relatively easy to build into user workflows and work incredibly well. What’s next? Look for greater adoption of AI-powered note-taking tools that can flag potential symptoms of specific conditions and automate coding and conditions that might be missed by providers.
- Improving access. At its core, AI and digital health technologies hold immense promise for expanding access to mental health services and addressing social determinants of well-being. While digital solutions may not be perfect substitutes for human-provided care, they can still markedly improve upon the limited or nonexistent services available to many, especially marginalized, populations. By streamlining triage, reducing biases, and breaking down stigma, AI-powered systems could significantly broaden the reach of mental healthcare and improve outcomes for those who have historically lacked access.
- Guide personalized care plans. AI will help analyze patient records (when available), personal demographic information, and vast amounts of treatment history to predict outcomes and recommend care pathways that will lead to better clinical outcomes. It will augment the existing note-taking solutions with better care plans.
- Improve training and feedback. Everyone can get better. AI will be able to listen to sessions and provide feedback to providers on how they can deliver better care, build better therapeutic alliances with their patients, and try new methods and techniques. This will not replace clinical supervision by more experienced clinicians but will reduce the cost and time involved and improve overall care. AI will also be able to provide simulated sessions so providers can practice on AIs to build skills with real patients. They can get more reps and practice on patients they may not be naturally well suited to treat.
- Automating intake and paperwork. AI can automate the gathering of information. Many care settings are required to perform elaborate intake and documentation before even beginning treatment for a patient. This can span multiple appointments before the patient seeking care can begin to get it. While regulatory changes could reduce this burden, AI can help step in and fill the gap by collecting information automatically from patients, asking follow-up questions and moving this information into the required forms. Intake will still involve people, but the overall time will be reduced, allowing reduced costs for most patients who can enter into care and more of our total dollars spent on care and less on administrative overhead.
- Automating talk therapy. AI tools can automate risk-assessment tasks in a clinical setting, and commercial apps that purport to automate talk therapy are an obvious next step for this technology. While this holds great promise for AI in behavioral health, beware of new digital consumer products that claim to do therapy leading to a “Wild West” of sorts in the Mental Health section of every device’s app store. Expect a regulatory backlash to follow. In the meantime, patients and providers will have difficulty discerning what works and what doesn’t.
- Digital Therapeutics. The FDA is approving digital therapeutics tools (i.e., Rejoyn, CBT-i Coach, Happify Health) at an increasing rate each year. The Centers for Medicare & Medicaid Services (CMS) added codes for Digital Mental Health Treatment in its 2025 Medicare Physician Fee Schedules. This means the prescription of digital mental health therapeutics can now be billed through Medicare and Medicaid, paving the way for private insurance to do the same. Much like your doctor writes you a prescription for medication or physical therapy, you can now be prescribed FDA-approved digital apps. Consumers ought to be aware of the likely differences between apps backed by clinical evidence, and apps that are available for purchase online. Marketing claims are not the same as peer-reviewed research.
The landscape for behavioral health tech innovations in 2025 isn’t all rosy. Delivering solutions that are truly patient-centered and have thoughtfully considered their impact on all the affected constituents is hard. Look for increasing confusion caused by the ease of creating apps that appear to be helpful but aren’t. Expect local, state, and federal laws to arise in response to protecting consumers from bad actors, which will have consequences for the builders of technology and those who need it.
Regulation might be necessary to protect consumers, but delaying helpful technology may create its own harm. The pace at which regulators act relative to the market is among the mysteries in store for 2025.
Loren Larsen is the CEO and co-founder of Videra Health, the leading AI-driven mental health assessment platform, and is a pioneer in leveraging video and artificial intelligence to assess and measure mental health. Prior to Videra, Larsen was the CTO of HireVue, a trailblazing video job interviewing platform with advanced machine learning algorithms. He also co-founded Nomi Health, a direct healthcare company striving to innovate within the healthcare service and technology space.