Hi friend,
I’m currently in San Francisco for the launch of our Hemingway SF Chapter. We hosted our official launch event last night and it was so inspiring to see the level of talent and energy dedicated to mental health innovation in this city. Thank you so much to everyone who made it!
So, please excuse the lateness of our round-up this week. We’ll be back to regularly scheduled programming from next Monday.
Hemingway SF Chapter Launch
Now let’s get into it!
What happened in mental health last week?
Business News:
⏺️ Beacon Biosignals upsized its Series B to $97M. The Boston-based company pairs an at-home EEG wearable with AI/ML analytics to generate objective neural biomarkers for CNS drug development. New investors include Palo Santo, Samsung Next, and JSL Health. (link)
⏺️ Click Therapeutics raised $50M in Series D funding and took over full commercialisation rights for CT-155. The prescription digital therapeutic for the negative symptoms of schizophrenia delivered a 62% relative improvement in symptom severity in its Phase III CONVOKE trial and holds FDA Breakthrough Device Designation. A day after the funding news, Click confirmed layoffs of some staff. (link)
⏺️ Kooth released their 2025 earnings, with revenue and profit down YoY. Revenue fell 5% to £63.3M (2.7% on constant currency), EBITDA dropped to £11.3M from £15.8M, largely driven by a spike in California direct marketing spend. The company remains profitable but a market cap of £38M on £63M of revenue reflects ongoing investor scepticism around California concentration risk. New contracts in New Jersey and Michigan ($2.6M year one) and growth in US Soluna registrations is the story investors will need more of before re-rating the stock. (link)
⏺️ Qualifacts acquired MethodOne, a controlled-medication dispensing platform built for opioid treatment programmes. The deal adds MAT/MOUD dispensing and inventory management directly into Qualifacts' behavioural health EHR platforms. There are ~2,224 Opioid Treatment Programs (OTPs) operating in the US, a segment growing at 5.6% CAGR since 2019. (link)
⏺️ CMS approved more than 150 companies for its ACCESS model, the new 10-year Medicare payment programme that ties reimbursement to clinical outcomes. Launching July 2026, it includes a dedicated behavioural health track covering depression and anxiety, with payment contingent on measurable PHQ-9 and GAD-7 improvement. Several mental health companies were included in the list, such as Headspace, Headway, SonderMind, Limbic, Jimini Health, SonderMind, and Slingshot AI. (link)
⏺️ Texas Children's Hospital received a $5M philanthropic gift to expand children's behavioural health services. The estate gift from Barbara LeGrange includes $1M in immediate funding with additional endowed support over time. (link)
⏺️ Talkiatry launched a browser-based psychiatric referral tool that embeds directly into cloud-based EMRs. Talkiatry Connect integrates with platforms including Athenahealth and eClinicalWorks, allowing primary care providers to make psychiatric referrals without leaving their existing workflow. (link)
⏺️ Spoony, an Australian mental health social platform, announced it will shut down by end of May. The app which launched in October 2024 and grew to 65,000 users with strong retention struggled to develop a revenue stream and raise funding. (link)
⏺️ Google updated Gemini's mental health crisis features and committed $30M to global crisis hotlines. The changes include a redesigned one-touch interface connecting users to crisis support and come weeks after the company was sued over a Gemini interaction linked to a user's death. (link)
⏺️ China's internet regulator published draft rules targeting AI-driven digital humans, with specific protections for minors. The rules would require disclosure when users are interacting with AI personas and restrict features designed to foster emotional dependency. (link)
Join the Hemingway Community
If you find these updates valuable and would like to connect with other leaders in mental health innovation, consider joining the Hemingway Community.
As a member of the community, you’ll join over 420 other mental health leaders and get access to exclusive content, events and resources.
Research News:
⏺️ LifeStance published outcomes data from nearly 180,000 patients showing improvements in both anxiety and depression. The analysis covered patients who initiated care between September 2024 and December 2025 across 33 states, using GAD-7 and PHQ-9 measurement tools. Results were consistent across all US geographic regions. (link)
⏺️ Two distinct major depression subtypes were identified based on brain white matter patterns, with researchers finding different treatment responses across each group. The machine learning study suggests AI could improve antidepressant selection which would be another step toward precision psychiatry for MDD. (link)
⏺️ A peer mentor programme embedded in the Kuamsha app — a digital depression intervention for adolescents — showed high feasibility and acceptability across sites in South Africa and Uganda. The programme achieved 68–100% recruitment rates and 89–100% training attendance, making it the first structured peer mentor model to be evaluated within a digital mental health intervention in low- and middle-income countries. (link)
⏺️ Researchers validated the DSM-5-TR Cross-Cutting Symptom Measure using data from 3,101 patients on Osmind's platform, identifying six clinically meaningful symptom domains. The findings provide practical guidance for interpreting this tool in real-world psychiatric practice as measurement-based care adoption grows. (link)
⏺️ A JMIR Mental Health paper argued that chatbot safety evaluations are misaligned with how psychiatric harm actually unfolds. Morrin, Auyeung, Agnew and colleagues proposed shifting from single-turn safety benchmarks to trajectory-based evaluation — tracking how risks like delusional drift, compulsive use, and suicidal crises accumulate across extended conversations rather than appearing at a single endpoint. They also called for safety research to incorporate proximal human outcomes following interactions, not just transcript analysis. All of these recommendations are completely reasonable. (link)
That’s all for this week.
Keep fighting the good fight!
Steve Duke
Founder of Hemingway
P.S. Find this valuable and want to support our work? Consider becoming a Hemingway Pro member.


