创新焦点：心理健康中的人工智能2025

EMERGING TECH RESEARCHInnovation Spotlight: AI in PitchBook Data, Inc. Nizar TarhuniExecutive Vice President ofResearch and Market Intelligence Mental Health Paul CondraGlobal Head of PrivateMarkets Research James UlanDirector of EmergingTechnology Research From algorithms to empathy, AI is taking a growing rolein addressing mental health needs Institutional Research Group Analysis PitchBook is a Morningstar company providing the most comprehensive, mostaccurate, and hard-to-find data for professionals doing business in the private markets. Aaron DeGagne, CFASenior Research Analyst,Healthcareaaron.degagne@pitchbook.com Data Innovation overview Collin AndersonSenior Data Analyst Mental health care remains deeply fragmented—marked by provider shortages,long wait times, and expensive care options. Against this backdrop, emergingapplications of AI are proving to be a meaningful lever to improve access to mentalhealth care and enable patient engagement between care sessions. pbinstitutionalresearch@pitchbook.com PublishingDesigned byChloe Ladwig There is significant untapped demand for patient-facing AI mental health solutions:Results from a recent survey show that half of large language model (LLM) userswith mental health needs use AI chatbots for mental health support.1This is aremarkable statistic, as these chatbots are not optimized to provide mental healthsupport, and shows that clinical-grade solutions need to evolve fast to catch upwith rapidly expanding consumer adoption of major consumer-facing LLMs suchas ChatGPT. There are core opportunities to provide care for those who are notcurrently engaged with professional care and those who need support betweensessions. It will also be important to promote AI mental health models safely giventhe risks involved in delivering subpar care. Published on June 3, 2025 Contents Innovation overview1Market analysis3Emerging providers4Deal activity5Appendix7 AI has the potential to influence the broader mental health care ecosystem onseveral fronts: matching patients with providers more effectively, unlockingnovel approaches to diagnostics, and reducing digital overhead for cliniciansthrough clinical documentation support. The AI mental health care startuplandscape includes a diverse set of solutions, which can be divided into severaldistinct categories: •Personalization:Platforms like Lyra Health and Spring Health use data tooptimize care matching and outcomes. •Provider support:Eleos Health, Limbic, and Ellipsis Health reducedocumentation overhead, enabling clinicians to focus more on patient care. •Diagnostics:Tools from acceXible and Kintsugi detect mental health changesfrom vocal and cognitive biomarkers. •Access:Consumer chatbots such as Wysa, Youper, and Sonar Mental Healthenable 24/7, stigma-free support through AI companions. AI-driven mental health startups operate within the broader digital healthecosystem, and beyond AI-first platforms, other digital health companies have mademeaningful strides in expanding access to mental health care. For example, digitalhealth startups like Headway and Grow Therapy connect patients with licensedtherapists, while condition-specific platforms, such as NOCD, focus on deliveringtargeted care for patients with specialized needs like OCD. These companies andothers are included in the teletherapy & behavioral health category of ouranalyst-curated healthtech verticaland tracked in our quarterlyhealthtechresearch. AI-supported mental health care, particularly on the patient-facing side, mayencounter resistance from professionals who argue that such tools fall short of thepersonalized care a licensed provider can offer. This critique is valid, as patientswould benefit most from direct, tailored support from clinicians. Still, the realityis that professional mental health services remain out of reach for many due to alack of access and high out-of-pocket costs. In this context, clinical-grade AI toolsoffer a preferable alternative to general-purpose chatbots that may lack scientificvalidation or oversight. A landmark study from Dartmouth recently found that theuse of mental health AI chatbots meaningfully reduced symptoms of depression,anxiety, and eating disorders.2 In our view, platforms that combine clinician-led services, such as patientengagement, with chatbot-style mental health support between sessions, are likelyto see the fastest adoption among providers. At the same time, access challengeswill continue to be addressed by an earlier generation of mental health tech startupsfocused on connecting patients to therapists and expanding insurance-covered care. One success story in this space is Sonar Mental Health, which has gained tractionnationally among school districts because they are often able to budget for theprogram using existing mental health grants.5 While there are significant market opportunities within the digital mental healthsubsector, this space is not without its challenges: AI cha