Morning, friends —
Happy 250, everyone. Grateful to have you all here as healthcare AI gets faster, weirder, louder, and more important. Let’s keep healing the world.
We’re also hiring a freelance reporter for Healthcare AI Guy. You’ll help us track healthcare AI news, find scoops, studies, launches, and drama, create exclusive reporting, and pull together the newsletter.
We’re looking for someone who is chronically online, scientifically literate, a strong writer, and deep in the AI arena.
If you’re interested, reply to this email with your CV, any writing samples, and a quick note on why you’d be a good fit. Now to the news:
Specialist models fire back
Anthropic gets into drug discovery
Primary care as AI's biggest opportunity
19 new tools/partnerships, 6 funding updates, new AI jobs & link-worthy content
Read time: 6 minutes
Our Picks ✨
Highlights if you’ve only got 2 minutes…
1/
Specialist models fire back
Just weeks after a widely shared Nature paper suggested that frontier models outperformed healthcare-specific tools, OpenEvidence published its own study arguing the opposite.
This time, the evaluation looked at 620 real questions physicians actually typed into OpenEvidence while caring for patients, rather than benchmark datasets. A blinded panel of 149 physicians across more than 30 specialties rated responses for accuracy, clinical utility, source quality, verifiability, and completeness. OpenEvidence came out on top across all five categories, outperforming Claude, Gemini, and GPT-5.5.
When the frontier models graded their own responses instead of physicians, GPT-5.5 preferred its own answers, while Claude and Gemini were much more aligned with human reviewers. It's another reminder that AI judging AI still has plenty of pitfalls.
This reinforces what we've heard repeatedly in the market: general models are incredibly capable, but healthcare is hard to replicate on a benchmark. Performance depends on clinical context, workflow, and evaluations that reflect how physicians actually practice medicine. It seems for now healthcare is still a specialization game. (link)(linkedin)
2/
Anthropic gets into drug discovery
Anthropic is making a bigger move into life sciences. Alongside the launch of Claude Science, an AI workbench for genomics, proteomics, and drug discovery, the company announced it's starting its own internal drug discovery program focused on neglected diseases that traditional pharma companies tend to overlook.
Anthropic says it isn't doing this to become a pharma company. It's doing it because building AI for drug discovery is a lot easier when you're actually trying to discover drugs yourself. The company wants firsthand experience with the workflows, bottlenecks, and decisions its customers face rather than building tools from the outside looking in.
It also fits a broader trend. Over the past few months we've seen OpenAI launch Rosalind for biology, Google-backed Isomorphic Labs push further into drug development, Meta-backed Biohub release new protein models, and now Anthropic stepping into the lab as well. The frontier AI labs aren't just selling models anymore. They're increasingly becoming practitioners in the industries they want to transform. It'll be interesting to see how much these model companies go direct vs enable incumbents. (link)(tweet)
3/
Primary care as AI's biggest opportunity
Everyone loves to talk about AI replacing doctors. A new report from Insight Partners argues the bigger opportunity is making primary care dramatically more capable. That's important because nearly half of US primary care physicians report burnout, patients wait more than a month on average for a new appointment, and doctors would need almost 27 hours a day to deliver all recommended care.
What's changing is the scope of what one PCP can do. Abridge and others are turning clinical conversations into documentation, coding, and decision support. Butterfly is putting AI-powered ultrasound into physicians' pockets. Mendaera is building robotic guidance so more clinicians can perform procedures that once required specialists. Tempus helps physicians match patients to treatments and clinical trials using multimodal data, while Cadence remotely monitors high-risk patients between visits to catch issues before they become hospitalizations. Tennr is tackling the referral bottleneck by automating patient intake and helping people get to the right specialist faster.
The broader scope is a much bigger opportunity than simply saving doctors a few minutes on notes. If AI can help primary care physicians diagnose more conditions, manage more patients, and keep people healthier before they need specialty care, that's one of the highest leverage bets in healthcare. (link)(linkedin)
Tools & Partnerships 🔧
Latest on business, consumer, and clinical healthcare AI tools and partnerships…
TOOLS
Anthropic launches Claude Science AI workbench: The company introduced a dedicated app for researchers that integrates over 60 scientific databases, manages computational workloads on demand, and generates auditable, publication-ready figures and manuscripts. (link)
Mayo Clinic approves 100+ clinical algorithms with AI: The health system deployed a new executive-led governance process in 2026 to review safety, workflows, and clinical parameters before letting healthcare workers use any tool. (link)
Microsoft launches Frontier Company for enterprise AI: The tech giant invested $2.5B in a new subsidiary that embeds 6,000 engineering and industry experts at customer sites to co-design and safely deploy enterprise-grade AI systems. (link)
Prediction model assesses risk of aneurysm rupture: An international team developed an AI tool that evaluates brain data over time to estimate the two year likelihood of an aneurysm bursting, outperforming standard clinical assessment scores. (link)
Youth turn to AI chatbots for mental health advice: A new cross-sectional survey found that roughly one in five US adolescents use AI conversational tools for mental health support, but nearly two-thirds never tell a parent or clinician. (link)
Mayo Clinic advances AI decision support systems: The healthcare organization hosted its AI Research Summit to highlight a shift from simple predictive tools to integrated decision-intelligence systems that actively guide clinical choices. (link)
Anthropic launches Claude Sonnet 5: The company unveiled its most agentic Sonnet model yet, built to autonomously navigate browsers and terminals, execute multi-step workflows, and close the performance gap with Opus-class models at a lower price point. (link)
Meta unveils non-invasive brain-reading AI: The tech giant introduced Brain2Qwerty v2, an end-to-end pipeline that uses fine-tuned LLMs to decode full natural sentences from non-invasive brain recordings in real time. (link)
Heidi Health launches Heidi Dictate: The clinical AI platform introduced a medical-grade voice-to-text tool that works across any desktop application, allowing clinicians to dictate notes, emails, and referrals directly via a single hotkey. (link)
Nourish ties AI-powered nutrition care to outcomes: Nourish launched a clinical outcomes model that links fees to patient results, using AI and data infrastructure to improve metabolic health outcomes and deliver value-based dietitian care. (link)
Evernorth unveils Pharmacy Forward program: The specialty pharmacy initiative leverages clinical data and AI-powered analytics to slash prescription processing times in half, boost medication adherence, and reduce clinical documentation workloads by 40%. (link)
Novant Health scales EHR-embedded AI to 19,000+ users: More than 19,000 team members now utilize AI tools integrated directly into the health system's Epic EHR, cutting administrative burden to actively combat clinician burnout. (link)
Study finds AI inherits human clinical biases: Mayo Clinic researchers found GPT-4 and GPT-5.1 frequently replicated human reasoning errors, including anchoring and confirmation bias, when interpreting oncology notes. (link)
Hallmark rolls out AI workforce operating system: The company launched an AI-enabled platform to give health systems real-time control over labor costs and incentives, aiming to optimize the industry's $900B fragmented spend. (link)
Hyro rolls out Care Intelligence layer: The agentic AI platform launched a comprehensive analytics and intelligence layer to help healthcare organizations measure, benchmark, and optimize patient access by extracting deep strategic insights directly from AI agent interactions. (link)
Zelis rolls out AI solution for dispute process: The healthcare technology company launched an AI-native tool designed to automate workflows and help payers navigate complex Independent Dispute Resolution processes under the No Surprises Act. (link)
Harvard trains pathology AI on 57M slides: Researchers developed a foundational AI model trained on a massive dataset of 57 million histopathology images to significantly improve diagnostic accuracy across multiple cancer types. (link)
PARTNERSHIPS
One Brooklyn Health + hellocare.ai: One Brooklyn Health selected hellocare.ai to deploy an AI-assisted virtual care platform across its system to support virtual nursing, patient safety, and continuous remote observation. (link)
Sevocity + Suki: Sevocity partnered with Suki to integrate an ambient AI assistant directly into its electronic health record platform to reduce documentation time for clinicians by 76. (link)
Deal Desk 💰
Spotlight on latest capital raises, M&A, and investments…
FUNDING
TJM Labs, a pharmacy AI platform automating operational work for pharmacies, raised $100M across three funding rounds in approximately six months led by Elephant, with participation from Arthur Ventures and Updata. The company also completed two acquisitions, EncoreRx and Pharmesol, expanding its automation and voice capabilities into a single pharmacy-native platform. (link)
Integral, the independent privacy layer for the real-world data economy, raised an $18M Series A with participation from Venrex, The General Partnership, Array Ventures, GreatPoint Ventures, and others. (link)
Queue, a Palo Alto, Calif.-based developer of a fully autonomous robotic pharmacy, raised $12.6M in seed funding. AlleyCorp led the round. (link)
Saltroad, a U.K.-based speech and language therapy startup, raised £1.5M led by Techstart Ventures and acquired AI documentation platform Ogma. (link)
M&A
OpenLoop + Hey Revia: OpenLoop acquired Hey Revia to add AI voice automation for prior authorizations, coverage verification, and pharmacy coordination to its Launchpad virtual care platform. (link)
Experity + Exdion Health: Experity acquired Exdion Health to expand its AI-powered RCM platform with automation for coding, charge capture, and claims workflows. (link)
market snapshot as of 7/5/26
Other Relevant News 🔍
News, podcasts, blogs, tweets, resources, etc…
👉 LAST CALL: Join our free live discussion to learn how physician groups are using agentic AI to improve patient access, grow revenue, & drive ROI
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…
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👇
