Listen to the episode on a streaming platform by clicking one of the links below:

• Listen on Spotify

• Listen on Apple Podcasts

• CMS walked back its initial proposal for a flat MA rate increase and instead finalized a 2.48% payment bump.

  • The increase between the initial and final proposal follows a familiar pattern and highlights the strength of the payer lobby. But the reprieve might only be temporary if CMS goes forward next year with a proposed update to their risk adjustment model.

  • Reduced supplemental benefits have slowed the growth of MA relative to traditional Medicare, but the Trump administration is exploring ways to continue increasing its share, including automatic enrollment.

  • Changes to the star rating system will boost payments to plans that succeed on core clinical quality metrics, but the removal of metrics related to appeal timeliness and outcomes could let payers get away with aggressive prior authorization tactics.

• A new study in JAMA finds that AI scribes reduce clinicians’ EHR times, but a claim that they also increase the volume of visits deserves careful scrutiny.

  • Reducing EHR time can also help providers avoid workforce burnout and boost retention, which might not be captured in existing measures of AI scribe ROI.

• Anthropic has developed a new AI model called Claude Mythos with capabilities that could significantly enhance cyberattacks. The company is withholding it from public release over safety concerns.

  • Anthropic is working with major tech companies like Amazon, Apple, Google, and Microsoft to harden their systems with the help of the unreleased model. The company estimates that rival firms could develop models with similar capabilities in as little as 6 months.

  • The fragmented and custom-built software that health systems rely on could be particularly vulnerable to AI-assisted cyberattacks, and health systems could find themselves paying large sums to security vendors to reduce the risk of devastating outages.

• At a recent public meeting, MedPAC presented study findings that the association between MA penetration and all-payer margins was consistently near zero across multiple sensitivity analyses.

  • Without more granular MA-specific data, the analysis may not isolate MA's true effect on health system margins—but that won't stop it from being held up as proof that MA pays enough.

  • Health systems could push federal policymakers to make better data available, following the example of states like California and Oregon.

Listen to the episode on a streaming platform by clicking one of the links below:

• Listen on Spotify

• Listen on Apple Podcasts

• Sutter Health and Allina Health have signed a letter of intent to form a combined system with 39 hospitals and 18K aligned physicians.

  • The 1,500-mile gap between the systems raises questions about systemness and governance, but there are opportunities for synergy in technology innovation, IT, RCM, and purchasing.

  • Diversifying across state lines to form a system with national scale can help diversify policy and market risks.

• On a panel at HIMSS26, CMS administrator Mehmet Oz and other administration officials expressed interest in introducing AI agents to Medicare that could help beneficiaries find providers or select plans.

  • Seniors might not trust AI, but that might not matter if these services and recommendations are pushed on them.

  • Unlike patients, AI navigator agents might not care so much about nonprofit health systems’ brand halo. We’ve raised similar concerns in the past about navigator services for commercial plans like UnitedHealthcare’s Surest.

  • Greater use of AI to detect Medicare fraud could root out bad actors, but false positives could create operational and reputational headaches for providers and health systems inadvertently caught in the crossfire.

• More than 75 strategy team members from 49 health systems joined us at last month’s Strategy Catalyst Summit in Arlington, Va. For this edition’s Key Market Dive, we’re digging deeper into four of our favorite insights and highlights from the sessions.

  • New AI deployments are making a visible impact on key strategic challenges like physician burnout and revenue capture, but they’re also demanding an increasing share of IT and transformation budgets. Strategy teams need to take on big picture questions about governance and vendor selection and can’t simply leave this work up to their IT teams.

  • In a hands-on demo showcasing how platforms like ChatGPT and Claude can be used for strategy work, we saw a wide range of AI fluency among participants. As these tools become more and more capable, strategy teams looking to transform their systems’ AI prowess would be wise to lead by example.

  • In another session, attendees discussed how reframing service line rationalization as a distribution problem can generate more organizational buy-in and sharpen thinking.

  • Stress testing different hypothetical market and policy scenarios can help strategy teams move from a reactive mindset to a more proactive approach in their strategic planning.

• This week’s featured graphic compares responses to THMA’s annual CXO priorities survey across a variety of executive roles.

  • Compared to other roles, CSOs are less focused on revenue capture and more focused on strengthening the workforce.

  • Executives across nearly every role are increasing their focus on AI.