Sponsored by: Lumeris
Nearly 100 million Americans lack access to primary care.
Tom by Lumeris helps close that gap. Tom is an AI-powered Primary Care as a Service platform that reviews patient panels, identifies the best next action, and executes it through proactive outreach, post-discharge follow-up, chronic disease management, and more.
👉 Book a demo today to see how Tom works.
Morning, freinds —
Let’s get right into it:
The 2026 Stanford AI index report
Biotech’s AI moment is here
AI's unintended consequences
24 new tools/partnerships, 9 funding updates, new AI jobs & link-worthy content
Read time: 5 minutes
Our Picks ✨
Highlights if you’ve only got 2 minutes…
1/
The 2026 Stanford AI index report
Per Stanford’s new 2026 AI Index Report, AI this year is defined by a few big shifts. Capability is accelerating fast (not plateauing), adoption is spreading across both enterprises and consumers at record speed. And importantly, the bottleneck is moving, from model architecture to data, workflows, and real world integration. AI is clearly and quickly becoming core infrastructure across industries. Nowhere is that more visible than in medicine:
Smaller models are winning in biology: Domain specific models with 100M to 200M parameters are outperforming models 100x larger on protein and genomics benchmarks, signaling a shift toward efficiency over brute force.
Virtual cells are emerging: Models like AlphaGenome and Evo 2 can simulate how cells respond to drugs and genetic changes, pushing more experimentation into software before the lab.
Data is the new bottleneck: Progress is now constrained by access to high quality biological data, with training sets expanding into the tens of millions.
AI is reshaping clinical workflows: Auto generated notes are cutting documentation time by up to 83%, reducing burnout and improving system efficiency.
Regulation is accelerating adoption: The FDA cleared 258 AI devices in 2025, most through faster pathways that do not require new trials.
Multi agent AI is outperforming doctors: Systems are hitting 85.5% accuracy on complex cases vs 20% for unaided physicians, with large gains over single model setups.
AI is now the front door to health information: Up to 92% of health searches trigger AI summaries, shaping how patients interpret symptoms and care decisions.
Ethics is expanding but still narrow: Discussion has doubled, but remains heavily focused on governance over accountability and equity.
Digital twins show early promise: In one diabetes trial, 71% of patients reached healthy glucose levels while reducing medication.
PS. We uploaded the 2026 Stanford AI Index report to our content hub 🥳
2/
Biotech’s AI moment is here
For decades, biotech has faced a frustrating paradox. We’ve mapped the genome, advanced biology into a data science, and built powerful tools, yet drug discovery has only gotten slower and more expensive. AI is starting to change that, acting as a connective layer across disciplines and helping scientists move faster through an increasingly complex system. This week was just another sign of this unlock being underway:
Amazon launches Bio Discovery: AWS introduced Amazon Bio Discovery, an end to end AI platform for drug design. It uses biological foundation models to generate and evaluate new molecules, and connects directly to lab partners like Ginkgo Bioworks and Twist Bioscience for synthesis and testing. Early users include major players like Bayer and Memorial Sloan Kettering, and the goal is to collapse timelines from discovery to validation. (link)
Novartis CEO joins Anthropic board: Vas Narasimhan joining Anthropic’s board is a notable signal of deeper integration between big pharma and frontier AI labs. Novartis has already been investing heavily in AI driven drug discovery, and this move gives Anthropic some customer lock in and strategic direction at the top. (link)
Novo Nordisk partners with OpenAI: Novo is deploying OpenAI’s models across its entire organization, from drug discovery to manufacturing and commercial operations. The partnership also focuses on upskilling Novo’s workforce, signaling AI is becoming embedded across the full pharma value chain vs. just research. (link)
OpenAI’s unveils life sciences research model: OpenAI introduced GPT-Rosalind, a frontier model built for biology, drug discovery & life sciences research. It shows strong gains in protein/chemical reasoning, genomics, and experimental design for scientific workflows. (link)
OpenAI shares real AI life sciences usage data: Roughly 200,000 people globally are already using ChatGPT each week for advanced life sciences work, primarily to synthesize and navigate scientific literature, with a growing share using it for analysis and research workflows. (link)
We uploaded the OpenAI report to our content hub 🥳
3/
AI's unintended consequences
AI was supposed to clean up one of healthcare’s biggest pain points: administrative complexity. Early evidence suggests it may be doing both things at once, making processes faster while also making the system more expensive.
A new PHTI report finds that AI is accelerating prior authorizations and billing workflows, but also increasing overall volume and intensity. Instead of reducing friction, some organizations are seeing more back and forth interactions, or “bot wars,” as payers and providers deploy competing automation.
On the billing side, AI tools like scribes are driving more detailed documentation and higher complexity coding, which leads to higher reimbursement and rising medical spend. At the same time, there is little evidence that AI is lowering the average cost per claim once implementation costs are included.
The underlying issue is incentives. AI is optimizing the system as it exists, not fixing it. Without changes to payment models, faster processes may simply mean faster, more expensive ones. We think this will get better overtime. (link)(linkedin)
TOGETHER WITH SUPERDIAL
If you care about AI in healthcare ops: SuperDial replaces manual payer calls with voice agents that actually get answers. 👉️ Click here to learn more!
Tools & Partnerships 🔧
Latest on business, consumer, and clinical healthcare AI tools and partnerships…
TOOLS
CMS expands ACCESS model with first cohort: CMS announced the first 150 organizations selected for its ACCESS model, which ties payments to measurable outcomes in chronic care and could become a key entry point for new players in Medicare. (link)
AI uncovers new GLP-1 side effects from social data: Researchers used LLMs to analyze 400K+ patient posts, identifying side effects of GLP-1 drugs not captured in trials. (link)
Anthropic releases Claude Opus 4.7: The latest model improves agentic coding performance and introduces new developer tools, while maintaining the same pricing as its predecessor. (link)
Autonomize launches healthcare AI platform V3: The platform combines a large clinical knowledge graph with pre-built agents and transparent decision-making to deploy healthcare workflows faster. (link)
Study analyzes 500K health chatbot queries: A Nature Health study found nearly 20% of conversations involve personal symptoms, with usage spiking at night and skewing toward personal concerns on mobile vs. professional use on desktop. (link)
Carrot launches AI platform for personalized fertility care: Carrot Intelligence uses a global clinical dataset to deliver context-driven guidance and improve care decisions across fertility and family health. (link)
Gemini adds mental health safeguards: Google introduced new crisis-response features in Gemini, including persistent hotline access during at-risk conversations and a $30M commitment to global mental health support. (link)
AI struggles with early diagnosis tasks: A Mass General Brigham study found chatbots failed at differential diagnosis over 80% of the time, though all LLMs achieved a correct final diagnosis more than 90% of the time and accuracy improved with more information. (link)
Mayo Clinic uses AI to improve heart risk prediction: AI analysis of cardiac imaging found that fat around the heart can significantly enhance long-term cardiovascular risk assessment. (link)
Atropos adds transparency to AI-generated evidence: A new review framework validates AI-driven clinical insights, following analysis of 33M treatment outcomes. (link)
CMS Medicare App Library goes live: CMS unveiled its first wave of vetted digital health and health AI apps, part of a broader push to build a connected healthcare ecosystem. (link)
Bunkerhill lands CMS reimbursement + FDA clearance: New billing code and first FDA-cleared AI to detect coronary and aortic valve calcium on routine contrast CT scans. (link)
Epic AI requires local validation: A meta-analysis found Epic’s AI models underperform out-of-the-box, reinforcing the need for health systems to validate tools before deployment. (link)
PARTNERSHIPS
Abridge + NEJM + JAMA: Abridge partnered with NEJM and JAMA to integrate trusted medical content into its clinical decision support platform for real-time use in clinician workflows. (link)
Wavelet Medical + Aegis Ventures: Wavelet Medical partnered with Aegis Ventures to develop an AI-powered noninvasive fetal EEG monitoring platform to improve detection of fetal distress and maternal outcomes. (link)
Optum Arcadia + Kivo Health: Optum Arcadia partnered with Kivo Health to deliver AI-powered virtual pulmonary rehab and chronic lung care at home, combining clinician-led care with remote monitoring and personalized treatment plans. (link)
Mount Sinai + SOPHiA GENETICS: Mount Sinai is adopting SOPHiA GENETICS’ AI platform to expand genomic testing and cancer research, streamlining variant analysis and lab workflows. (link)
Monument Health + Epic: Monument Health is using generative AI tools within Epic to support clinical workflows, with adoption rates exceeding national benchmarks. (link)
Houston Methodist + Midstream Health: Houston Methodist deployed Midstream Health’s AI financial platform across enterprise operations. (link)
Premier Health + Wolters Kluwer: Premier Health adopted UpToDate Expert AI systemwide to support clinical decision-making. (link)
Southwest General + Oracle: Southwest General deployed Oracle’s clinical AI agent to reduce physician documentation time. (link)
Deal Desk 💰
Spotlight on latest capital raises, M&A, and investments…
FUNDING
Alloy Therapeutics, an AI-driven biotechnology infrastructure company, raised $40M in a Series E at a $1.0B valuation led by 8VC and JIC Venture Growth Investments. (link)
Joyful Health, an AI financial infrastructure company for healthcare revenue operations, raised a $17M Series A led by CRV, bringing the company's total funding to $22M. (link)
Keebler Health, an AI-powered risk adjustment platform, raised $16M in series A funding. Flare Capital Partners led. (link)
ScreenPoint Medical, an AI solution for breast cancer detection and risk assessment, raised $14M in funding from Insight Partners and Siemens Healthineers. (link)
Coral, an NYC-based healthcare automation platform, raised $12.5M in seed funding. Lightspeed and Z47 led the round. (link)
Helical, a virtual AI platform for drug discovery, raised $10M in seed funding. Redalpine led. (link)
Ethermed, a Philadelphia, PA-based company providing AI-driven automation for prior authorization and medical-necessity workflows, raised $8.5M in Series A funding. Enfield Capital Partners and Blue Marlin Partners led. (link)
Foresight Health, building an AI-native healthcare provider, raised $8.5M in seed funding from Quiet Capital, Susa Ventures, and Neo. (link)
ONTO Health, an AI-enabled, physician-led fertility and longevity provider, raised a $20M Series A from ARTIS and Humania. (link)
GRANTS
Google + Johnson & Johnson Foundation: Google.org and the J&J Foundation launched a $10M initiative to train rural healthcare workers in AI to reduce administrative burden and support adoption of new tools. (link)
market snapshot as of 4/17/26
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…
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👇
