McKinsey's $970M savings estimate that's fueling the payer-provider AI arms race

🌟 Editor's Note

HIMSS 2026 marked the vibe shift — buyers stopped asking "what can AI do?" and started demanding "prove it works." The real story is Mass General Brigham building governance for shadow AI users and Children's Nebraska deploying agentic systems that solve actual problems instead of running pilots. Meanwhile, the bot-versus-bot arms race between hospitals and insurers is generating billions in AI spending but zero systemic improvement. The unsexy winners are building infrastructure, not chasing innovation theater.

🎙️ THE BIG STORY

AI Agents take over HIMSS 2026 this week

HIMSS was different this year. The AI demos took a back seat to something far more valuable: proof.

Greg Samios, CEO of Wolters Kluwer Health, told Newsweek that while it's easy to spin up an AI product quickly, demonstrating value takes considerable time. Translation: The honeymoon's over. Healthcare's moving from "look what AI can do" to "prove it works."

The conversation on the floor wasn't about features. It was about roadmaps, five-year plans, and governance frameworks.

Seth Hain, Epic's SVP of R&D, recalled that just two years ago at their User Group Meeting, people questioned whether their AI demonstrations were even real. Now Emmie, their AI patient assistant, is live at Sutter Health.

From ambient scribes to autonomous agents in under 24 months. The technology moved fast. Now the implementation has to catch up.

💰 THE AI ARMS RACE

McKinsey's $970M savings estimate that's fueling the payer-provider AI arms race

Hospitals and insurers are locked in an AI escalation that's reshaping healthcare's financial foundation. McKinsey estimates that for every $10 billion in revenue, AI could save insurers $970 million through claims management, medical prior authorization requests, and guiding clinical care.

Meanwhile, HCA Healthcare expects about $400 million in 2026 cost savings from AI initiatives, using automation for revenue management. HCA's CFO previously described this as a response to "growing denial and underpayment activities from payers."

The stakes: A Blue Cross Blue Shield analysis found that roughly $663 million in inpatient spending and at least $1.67 billion in outpatient spending may be tied to more aggressive, AI-enabled coding practices nationwide.

Healthcare AI spending reached $1.4 billion in 2025, nearly triple 2024 levels. Health systems accounted for roughly $1 billion (75%), while payers invested only about $50 million.

UnitedHealth Group projects AI could save it nearly $1 billion in 2026, planning to invest $1.5 billion in AI this year. Smaller rival Humana estimates over $100 million in savings over a few years.

The irony? Christina Silcox of Duke-Margolis Institute observed that bot-versus-bot competition is "intrinsically a situation where no one's going to win."

🏥 ENTERPRISE PLAYBOOK

Mass General Brigham's Reality Check on AI Governance

While vendors sell AI dreams, Mass General Brigham is building the unsexy infrastructure that actually makes AI work at scale.

Jane Moran, MGB's chief information and digital officer, told the HIMSS26 AI in Healthcare Forum that AI is advancing rapidly, but healthcare is approaching it with caution because of the high stakes involved.

The shadow AI problem: MGB recognized that many staff were already using AI tools on their own — tools that were not always compliant. Researchers were using public-facing consumer AI tools for cancer and Alzheimer's research, lung cancer imaging, and more.

Their solution: Mass General Brigham stood up a "secure, multi-model prompt platform" called the AI Zone.

Governance evolution: As recently as 2024, MGB relied on its existing governance framework for AI. By 2025, they recognized the need for a formal AI governance process.

Moran's philosophy: "Safely and effectively scaling AI depends on people and process as much as actual models and technology. It's about building a sustainable, human-centered AI capability that drives meaningful outcomes across the enterprise."

Key risks: patient safety, bias in data, cyber exposure from sensitive data access, and governance challenges around accountability and validation.

🔬 AGENTIC AI IN ACTION

Children's Nebraska: From Pilots to Production

Ryan M. Cameron, Children's Nebraska's newly appointed Chief Information and Innovation Officer, has a blunt message: Stop running pilots. Start solving problems.

Cameron believes agentic AI is a game-changer. "Children's Nebraska is deploying agentic and workflow-aware AI in operational environments and evaluating the best, safest, most reliable areas for deployments in clinical environments."

The key principle: "Pursue a small set of agentic AI projects, not as pilots for innovation's sake, but as solutions to concrete problems."

Case in point — HeyMedicaid.med: Children's Nebraska built this platform using agentic AI to proactively manage Medicaid eligibility, enrollment, and renewal workflows. It addresses one of healthcare's most persistent failures: children losing coverage not because they're ineligible, but because the process is too complex to navigate.

The system keeps a "human in the loop" while proactively identifying eligibility risk, guiding families through enrollment, and coordinating across hospital, state, and community partners.

Cameron noted: "Last year, healthcare organizations reported most AI pilots failed to deliver the ROI projected. You have to start with the problem first and then build the tech around it."

His advice to C-suite: "Reframe AI and automation as strategic infrastructure, not experimental tools. Move beyond pilots toward scaled, governed deployments paired with enterprise priorities like access, patient satisfaction, clinician retention, and financial resilience."

🎙️ THE TAKEAWAY

HIMSS 2026 marked healthcare AI's transition from proof-of-concept to proof-of-value. The bot-versus-bot escalation between hospitals and insurers reveals fundamental misalignment — both sides deploying AI to win a zero-sum game rather than redesign broken processes.

Mass General Brigham's governance-first approach and Children's Nebraska's problem-first deployment strategy represent operational maturity over innovation theater. They understand that AI's value isn't in automation velocity — it's in sustainable transformation that improves outcomes while managing risk.

The strategic separation is clear: Organizations treating AI as experimental innovation will chase pilots indefinitely. Organizations treating AI as infrastructure are already embedding agents into production workflows, building governance frameworks, and measuring concrete outcomes.

Epic's 85% adoption rate settled whether AI works in healthcare. The question is whether your organization is building the people, processes, and governance infrastructure to scale it.

Talk soon— Ram Menon, CEO of Avaamo