Good morning, everyone —
How are provider groups actually turning AI from pilots into real P&L impact?
A lot of the AI adoption conversation has focused on large health systems. We wanted to go deeper and shine a light on mid-market outpatient physician groups (the majority of care delivery) and how they're implementing agentic AI across front- and back-office workflows to improve patient access, grow revenue, and drive ROI.
Join us for a free live discussion on July 9th featuring top leaders from the front lines: an AI operating advisor, an AI founder, & a physician group executive actively implementing these technologies at scale.
Now, to the health x AI news:
Midjourney's wild pivot into healthcare
AI beats doctors... sort of
Power to the people
23 new tools/partnerships, 11 funding updates, new AI jobs & link-worthy content
Read time: 6 minutes
Our Picks ✨
Highlights if you’ve only got 2 minutes…
1/
Midjourney's wild pivot into healthcare
In one of the most unexpected product launches of the year, Midjourney, the AI image generation company, announced a new division called Midjourney Medical alongside a futuristic full-body scanner that feels more science fiction than healthcare.
The scanner uses hundreds of thousands of ultrasound transducers arranged in rings around the body to capture a full-body scan in roughly 60 seconds, all without radiation, magnets, or contrast agents. Midjourney says the system could eventually deliver MRI-like imaging at a fraction of the cost while being safe enough for routine, repeated scans. The hardware is being developed in an exclusive partnership with Butterfly Network ($BFLY), which has a patented specialized chip that powers this ultrasound technology. The stock was up 50% on the news.
The vision goes far beyond the scanner since Midjourney plans to open its first Midjourney Spa in San Francisco's Union Square in 2027, combining full-body scans with saunas, cold plunges, hot tubs, etc. The company ultimately wants to deploy 50,000 scanners globally, creating a new preventative health infrastructure.
Midjourney’s founder, David Holz, is also fascinating and isn't a traditional healthcare entrepreneur. Before Midjourney, he founded Leap Motion, one of the earliest companies working on hand tracking and spatial computing. Midjourney itself has famously broken many Silicon Valley rules, bootstrapping without venture capital and growing into a massively profitable business through its image generation platform (where the first Healthcare AI Guy avatar was created). That success is now funding ambitious bets like this.
There are still some unanswered questions (ex., regulatory approval, imaging performance claims, etc) but the broader idea is super compelling.
Imagine combining a full-body scan with your lab results, wearable data, medical history, and AI-powered analysis. Instead of interacting with healthcare only when something goes wrong, you could have a continuously updated picture of your health over time.
After a decade of talking about preventative care, this is one of the best product launches that genuinely feels like it was designed for that world. The future should look like the future, and this is about as futuristic as healthcare has gotten. (link)(linkedin)
2/
AI beats doctors... sort of
Two new Nature studies generated plenty of headlines this week after finding that AI agents outperformed physicians on emergency medicine and care management tasks. But the details are a bit more nuanced than the headlines suggest.
In the first study, an AI agent called MIRA achieved 87.8% diagnostic accuracy across 574 emergency department cases, outperforming physicians at 78.1%. A second study found Google's AMIE produced higher-rated care management plans than 21 primary care physicians across 100 longitudinal patient cases.
That said, both systems operated in highly controlled environments. MIRA worked inside a sandboxed EHR and AMIE managed idealized patient scenarios without prior authorizations, formulary restrictions, scheduling delays, or the countless real-world variables that make medicine messy.
Still, the results matter and suggest AI systems are starting to actually execute workflows and care plans vs just helping answer questions. That's a much bigger shift. And as Eric Topol pointed out, today's models are already more capable than the ones used in either study.
3/
Power to the people
One of the most interesting healthcare trends isn't happening inside hospitals. It's happening in the hands of patients.
Last week, AI researcher Amy Deng shared in a post that went viral how she used AI to help untangle a persistent health issue. Instead of relying on a single appointment, she tracked a dozen health metrics, ran extensive bloodwork, analyzed the results with AI, and continuously tested different interventions. The result was a combination of factors including iron deficiency, under-fueling, hormone irregularities, and medication dosing issues that had been impacting her energy levels.
Similarly, a Columbia ER physician wrote in the New York Times that ChatGPT provided more personalized guidance on her own lab results than her doctor did. While her physician's recommendation was to schedule another appointment, ChatGPT immediately analyzed her results, discussed lifestyle factors, and suggested a customized action plan.
Neither story suggests AI should replace clinicians and risks remain but it highlights an important trend: patients now have access to tools that can help them better understand their own data, ask better questions, and play a more active role in their care than before. AI is turning patients into much more informed and engaged participants in the healthcare process, and we’re here for it. (tweet)
Tools & Partnerships 🔧
Latest on business, consumer, and clinical healthcare AI tools and partnerships…
TOOLS
UnitedHealth expands AI-powered patient scheduling: The insurer is using AI voice agents to call physician offices and schedule appointments on behalf of members, extending its automation efforts across care coordination. (link)
Anthropic and U.S. officials work to restore model access: Anthropic is in discussions with the Trump administration to resolve restrictions placed on its Fable 5 and Mythos 5 models following concerns about safeguard vulnerabilities and misuse risks. (link)
OpenAI brings frontier health AI to free users: OpenAI says GPT-5.5 Instant has reached frontier-level healthcare performance, improving HealthBench scores, outperforming physicians across five evaluations, and reducing potential factuality errors in health conversations by 71% over the past two months. (link)
Epic expands its long-term AI strategy: New reporting highlights Epic’s push to embed AI deeper into healthcare operations, starting with scheduling and administrative workflows as part of a broader platform strategy. (link)
FDA clears record number of AI medical devices: The agency has now authorized 1,524 AI-enabled medical devices since 1995, including 92 new approvals in Q1 2026 alone, reflecting accelerating adoption across healthcare. (link)
New benchmark exposes clinical AI gap: Mass General Brigham researchers developed BRIDGE, a multilingual benchmark showing that while leading AI models score above 90% on medical licensing exams, their performance drops sharply on real-world clinical text and patient-care tasks. (link)
OpenAI model helps solve rare-disease cases: Boston Children’s and Harvard researchers used o3 Deep Research to revisit 376 unsolved pediatric genetic cases, surfacing leads that doctors confirmed into 18 new diagnoses. (link)
Study finds AI-related errors drive more lawsuits: University of Cambridge researchers found adverse events involving AI significantly increase blame toward hospitals and the likelihood of legal action, particularly when physicians are not actively involved in oversight. (link)
Anthropic showcases healthcare adoption of Claude: New customer case studies highlight Epic, Banner Health, and Carta Healthcare using Claude to accelerate product development, reduce clinician administrative work, and streamline clinical data processing. (link)
AI mental health tools face growing scrutiny: A JAMA Psychiatry review highlighted both the promise and risks of AI in mental healthcare, emphasizing concerns around unpredictable performance and patient safety. (link)
Ambient AI scribes may change care patterns: A study of 20,000+ primary care visits found AI scribes captured more neuropsychiatric symptoms than human documentation, but clinicians were less likely to diagnose or treat depression during those encounters. (link)
States turn to AI to manage Medicaid: More states are deploying AI for eligibility checks, chatbots, and administrative workflows in Medicaid programs, raising questions about oversight and the risk of wrongful benefit denials. (link)
Lumeris adds AI symptom checker to Tom platform: The new capability lets high-risk patients report symptoms and ask health questions between visits, helping care teams identify emerging issues earlier. (link)
PARTNERSHIPS
WHOOP + HealthEx: WHOOP partnered with HealthEx to connect members’ EHR data directly into the WHOOP platform, combining clinical records with wearable insights for more personalized health guidance. (link)
Takeda + Boltz: Takeda partnered with Boltz to deploy biomolecular AI foundation models across its research organization to support drug discovery and biologics development. (link)
Everlab + Decagon: Everlab selected Decagon to power AI-driven customer operations as it scales its preventative healthcare platform. (link)
Sword Health + Portugal National Health Service: Sword partnered with Portugal’s National Health Service to make its AI-powered musculoskeletal care platform available nationwide through the public health system. (link)
CodaMetrix + 5 Health Systems: CodaMetrix partnered with five health systems to launch the Coding Quality Council, a governance initiative focused on establishing standards for AI-powered medical coding quality. (link)
Deal Desk 💰
Spotlight on latest capital raises, M&A, and investments…
FUNDING
Connie, an AI-enabled rollup of Medicare Advantage brokerages, raised $40M. HealthQuest Capital led. (link)
Telepatia, an AI-native clinical platform for Latin America, raised a $33M Series A led by a16z. (link)
Prosper AI, an NYC-based AI platform designed to combine patient scheduling, insurance verification, and patient billing, raised $30M in Series A funding. a16z led the round and was joined by Base10 and existing investors Emergence Capital, Y Combinator, and Company Ventures. (link)
Gero, a Singapore and San Francisco-based ‘physics‑first’ AI drug discovery company focused on age‑related diseases and the biology of aging, raised $17M in funding. Melnichek Investments led the round and was joined by others. (link)
Rocapine, a Paris, France-based AI-native wellness venture studio, raised $13M in Series A funding. Educapital led the round. (link)
Clair Health, a San Francisco-based developer of a noninvasive continuous hormone monitor for women, raised $11.6M in seed funding. Khosla Ventures led the round and was joined by a16z speedrun, Brydge Club, Treehub, and others. (link)
Frontier Health, a developer of AI agents for the NHS, raised £9.7M led by Atomico. (link)
Kimba, an AI sleep tech company, raised $6.5M in funding led by Selva Ventures. (link)
Vali Health, an SF-based maker of AI agents for home care agencies, raised $6M in seed funding from Bonfire Ventures, Supernode, and Comma Capital. (link)
Kinomatic, an AI and VR-driven orthopedic surgery planning and concierge recovery company, raised $4M in seed funding. Waterline Ventures led. (link)
M&A
Thoreau + Ensemble: Thoreau made a strategic investment in Ensemble at a roughly $12B valuation, positioning the revenue cycle leader as a platform for future AI-enabled healthcare administration acquisitions and expansion. (link)
Other Relevant News 🔍
News, podcasts, blogs, tweets, resources, etc…
AI Job Opportunities 💼
Explore our AI Job Board or contact us to feature roles in our newsletter…
Visuals of the Week 📸
Funny memes, cool pics, and interesting data from around the web…
midjounrey founder/ceo
That’s it for this week friends! Back to reading — I’ll see you next week.
Stay classy,
PS. I write this newsletter for you. So if you have any suggestions or questions, feel free to reply to this email and let me know
How was this week's newsletter? Tap your choice below👇
