AI-in-the-Loop and the Next Chapter of Payer Performance

By Akros Advisory in collaboration with HTEC Group

The New Frontier for Payers

At HLTH 2025, Akros Advisory and HTEC Group convened two closed-door lunches for payer, provider, and technology leaders. One explored Pain Points to Performance; the other unpacked AI, Quality, and Care Management.

Across both rooms, the message was clear: Payers are no longer just financial intermediaries. They are becoming system designers.

This shift requires a new mindset. Technology isn’t just automating tasks anymore. It’s amplifying what people can do by making decisions sharper, data smarter, and care more connected.

1. The Payer Imperative: From Pain Points to Patterns

Every health plan executive can list the pain points: interoperability, poor member experience, manual processes, rising costs. What emerged at HLTH was not a new list of problems, but a new pattern of thinking:

• Interoperability before innovation. The data problem is behavioral, not technical. Until organizations treat information as a shared public good rather than a proprietary asset, AI will only ever be as smart as the walls that contain it.

• Designing for empathy. Efficiency cannot come at the cost of emotional presence. Technology should make care more human, not less.

• Reframing ROI. Cost reduction is not the endpoint; rather, it’s the byproduct of systems designed to make humans better at their jobs.

One of the participants noted:

“Everyone wants AI, but no one can explain what problem it’s supposed to solve. The real opportunity is in teaching teams to think differently.”

2. “AI in the Loop”: A Design Philosophy, Not a Tool

A breakthrough moment in the conversation reframed the industry’s favorite buzzword: Instead of “human in the loop,” leaders called for “AI in the human loop.”

It’s a subtle but powerful inversion. “Human in the loop” assumes AI leads and humans intervene. “AI in the loop” assumes humans lead with AI woven throughout the system.

That distinction matters for care management, Stars improvement, and beyond. In an “AI in the loop” model, technology amplifies human intent, surfacing insight, reducing friction, and contextualizing care, while people retain agency and empathy.

It’s a design philosophy that determines whether technology amplifies or erodes humanity.

3. Data Integrity Is the New Clinical Quality

Interoperability may be mandated, but it isn’t yet meaningful.

Most payers have met the letter of compliance without achieving the spirit of empowerment. True modernization will come when interoperability stops being about file formats and starts being about fluid trust.

The group drew parallels between blockchain’s unfulfilled promise and the current state of health data exchange: both were built for openness, yet both became siloed.

To deliver on value-based care, data can no longer be a competitive moat. It must become an ecosystem asset by being transparent, auditable, and used to improve outcomes, not just measure them.

4. Empathy as a Strategic Advantage

As the group discussed the implications of AI-driven quality measurement, a quiet consensus emerged: Empathy is the payer’s most underutilized performance lever.

AI can triage, summarize, and guide but it must never diagnose in isolation. It can inform the care conversation, but it cannot replace the emotional presence that drives trust and adherence.

When designed thoughtfully, AI becomes a partner in compassion by surfacing insights from behavioral data, wearables, and claims so that care managers can connect earlier and more meaningfully.

The system doesn’t need fewer humans. It needs more humane systems.

5. The Role of Regulation and Responsibility

Participants debated the widening gap between Europe’s overregulation and America’s underregulation of AI.

The EU’s AI Act, requiring re-certification for every algorithmic update, risks strangling innovation through bureaucracy. The U.S., meanwhile, risks accelerating innovation without ethical oversight.

Both extremes fail the same test of governance without guidance.

What payers need instead is principled agility, frameworks that protect patients while enabling experimentation. That balance will define which organizations earn trust in an increasingly automated world.

6. Culture: The Hardest System to Modernize

Technology moves fast. People often don’t.

The most forward-thinking payers at HLTH spoke not about cloud migrations or LLM integrations, but about cultural readiness and the courage to test, fail, and learn.

AI maturity will not be measured in models deployed, but in mindsets changed. Plans that invest in AI literacy, teaching teams to prompt, question, and think with machines, will outpace those that only invest in tools.

One leader said it best:

“AI isn’t coming for healthcare. It’s coming with it. The question is whether we’ll meet it with curiosity or fear.”

7. The Future of Payer Performance

The payers shaping the next decade will be those who:

• Redesign performance around people, not processes.

• Use data to empower, not entrench.

• Adopt AI as augmentation, not automation.

• Treat trust as infrastructure, not intention.

The work ahead is not about choosing between technology and humanity. It’s about designing systems where they can coexist, scale, and strengthen each other.

Until that equation changes, until data is shared, design is human, and intelligence is collective, healthcare AI will only ever reflect our fragmentation, not our potential.

Closing Thought

Healthcare’s future won’t be defined by who builds the fastest model. It will be defined by who builds the most thoughtful system, one that remembers that care, at its core, is still human.

And that is exactly where payer leadership begins.