Report from the Field: Six Signals Reveal the Future of Payment Integrity

Payment integrity leaders from the country’s top health plans gathered at the 2026 Power of Payment Integrity Conference. They found a community ready to take on today’s challenges and create tomorrow’s solutions.

If there was a single theme that cut across this year’s conversations at the annual Power of Payment Integrity Conference, it was this:

Payment integrity can no longer hide in the back office. It has become a central lever in how healthcare is managed, financed, and ultimately transformed.

“The old model is changing. The shift in PI is happening very quickly —especially for our industry — and it’s exciting.”
-Susan Luden, Highmark

This evolution showed up both from the stage and in the audience. Across 24 sessions spanning two days in Nashville, what leaders were sharing and planning for went beyond incremental improvements. They were describing a function that is expanding in scope, increasing in strategic importance, and, in many cases, being fundamentally redefined.

This report highlights six signals from the field that point to where payment integrity is heading next.

1. CMS is accelerating innovation in payment accuracy — and raising the bar for everyone else

A notable shift across conversations was the growing recognition that innovation in payment integrity is no longer being driven primarily by commercial plans. It’s coming from Centers for Medicare & Medicaid Services (CMS).

From mandates around interoperability (FHIR APIs, payer-to-payer exchange) to increasing expectations for data transparency and prior authorization digitization, CMS is actively reshaping the infrastructure of how payment accuracy is achieved. And in several respects, it’s moving faster than the commercial market.

“Healthcare costs are exploding, the trust fund clock is ticking, and we need to get these bad actors out of the program.”
— Jennifer Dupee, CMS

This creates a new dynamic:

• CMS is defining the technical and operational baseline (data access, exchange standards, real-time workflows)
• Commercial plans are being forced to adapt within legacy systems and fragmented architectures
• The gap between what is possible and what is currently operational is becoming more visible

At the same time, CMS modernization efforts — including exploration of next-generation claims platforms to replace aging infrastructure — signal that even the largest, most complex payer in the system is investing in foundational change.

The implication for payment integrity leaders is significant: The future state of PI is increasingly being shaped outside the four walls of the health plan.

“Government payers are out in front, and that has historically never happened. The pace of change is happening faster than ever before, and technology is a big part of that.”
— Christina Hedge, Booz Allen Hamilton

Whether it’s real-time data exchange, embedded workflows within provider systems, or more transparent payment logic, the direction is being set at the ecosystem level — not plan by plan.

For commercial organizations, this raises a strategic question:

• Will you move in step with CMS — adapting your PI program to align with where the system is going?
• Or will you continue optimizing within current constraints, and risk falling behind as the infrastructure evolves around you?

Key takeaway: Payment integrity performance will depend not just on what you build internally — but on how well you integrate into the external systems being standardized around you. 

2. AI in payment integrity is moving past experimentation — and into operational reality

The conversation around AI in payment integrity hasn’t quieted since we met in this forum last year — it has matured. But what was once dominated by experimentation is now centered on operationalization, governance, and measurable value.

Brenton Hill, COO of CHAI, captured the shift clearly:

“It’s not just AI that’s going to move the industry forward — it’s the controls and governance around those that are going to do that more favorably for us.”

That shift is being driven in part by external pressure. Regulatory scrutiny is accelerating — from CMS legal challenges to growing state legislation — and NCQA will require formal AI governance attestation by 2027. AI is no longer a sandbox; it’s becoming a regulated operating environment.

As a result, leading organizations are focusing less on tools and more on provable controls: defined use cases, model transparency, performance monitoring, and audit trails that can explain exactly how decisions are made. The concept of “human in the loop” is also evolving toward “intentional human checkpoints” — designed friction that ensures real oversight, not passive review.

“No adverse determination should be made automatically by AI.”
— Andrew Berry, Humana

At the same time, organizations are working to break out of “pilot-itis.” As Andrew Berry of Humana emphasized, AI must be treated as a strategic investment, with clear business objectives, success metrics, and a path to production — not isolated experiments with limited impact.

Importantly, expectations are becoming more grounded. AI is not replacing payment integrity teams — it’s making them more effective. Its strongest use cases today are narrowing focus: reducing false positives, improving claim selection, and surfacing evidence for human review.

Key takeaway: The next phase of AI in payment integrity will be defined by disciplined execution rather than bold experimentation.

3. Programs that manage at the concept level are outperforming — and setting a new standard

One of the most concrete signals of maturity came from benchmarking data presented by Tom Baggett.

Across 18 health plans, $49.1 billion in paid amounts were reviewed, yielding $13.4 billion in recoveries — a 29% recovery rate. But the more important insight was how that value was distributed.

The top 10 concepts accounted for ~$885 million in savings. The remaining ~1,000 concepts accounted for $12.5 billion. This long-tail value is where high-performing programs are separating themselves. They are:

• Managing PI work at the concept level, not just audit volume
• Expanding concept coverage while simultaneously improving yield
• Using structured frameworks (heatmaps, quadrant models) to guide resource allocation
• Treating ideation as an analytical discipline, not an administrative task

Perhaps most importantly, they’ve embraced a different mindset: More visibility is better — even when it reveals more problems.

“This is a roadmap. This is not a scorecard. Knowing what we are doing today is great — but knowing what we are not doing is better.”
— Tom Baggett, ClarisHealth

As Baggett put it, the goal is a “better set of problems.” Programs that surface more leakage, more complexity, and more opportunity are operating with greater clarity.

This also reinforces a critical operational truth: Stale concepts are one of the biggest sources of missed savings. High-performing teams maintain a high-velocity concept pipeline, reviewing and refreshing regularly. And retiring underperforming concepts is just as important as creating new ones.

Key takeaway: Granularity drives performance. Programs that understand their work at the concept level — and actively manage it — outperform those that don’t.

4. Provider AI adoption is accelerating — and changing the nature of the problem

Another major shift is coming from the provider side.

As highlighted by Aurene Wilford, an AI arms race is already underway. Providers are rapidly adopting AI tools to optimize documentation and coding — increasing what is often referred to as “coding intensity.”

At the same time:

• 80% of payers and 70% of providers are actively building AI strategies
• AI is moving from experimentation into production workflows
• Adoption is accelerating at the functional level, even if not yet enterprise-wide

The impact on payment integrity is significant. More sophisticated documentation and coding practices mean:

• Traditional rules and static edits become less effective
• Detection becomes more complex and nuanced
• The volume of “technically correct but economically aggressive” claims increases

“Providers said: ‘Operational complexity is rising faster than we’re being reimbursed. Denials are no longer exceptions — they’re part of our workflow.’”
— Aurene Wilford, KLAS 

This doesn’t mean the payer-provider relationship is deteriorating. In fact, Wilford noted that collaboration is improving in many areas. Shared dashboards, joint governance, and value-based care initiatives are all gaining traction.

However, the nature of PI work is changing. It’s no longer just about identifying clear errors. Increasingly, it’s about:

• Interpreting intent within more complex documentation
• Adapting quickly to evolving billing behaviors
• Ensuring that AI-enabled optimization on one side is matched by equally sophisticated oversight on the other

Key takeaway: As providers get smarter with AI, PI programs must evolve just as quickly — or risk falling behind.

5. “Shifting left” is no longer just about prepay — it’s moving to pre-submission

For years, the industry’s north star has been clear: shift left from post-pay recovery to prepay prevention. That goal is already being achieved in many organizations. So the next frontier is emerging beyond it.

“My favorite phrase is ‘real-time.’ Even prepay takes too long!”
— Dr. Priscilla Alfaro, Blue Cross NC 

As both Scott Magit and Priscilla Alfaro emphasized, the conversation is moving from prepay to pre-submission — intervening before the claim is even sent. This is a fundamentally different model.

Instead of:

• Denying claims
• Requesting records
• Managing disputes and reconsiderations
• Reprocessing corrected submissions

The goal becomes: Help the provider get it right the first time. The benefits compound quickly:

• Fewer denials
• Fewer disputes and reconsiderations
• Reduced administrative burden for both payers and providers
• Faster, cleaner payment cycles
• Improved provider relationships

“Keeping these edits and audits behind four walls and not giving providers the visibility into it — I think there’s diminishing value in that.”
— Scott Magit, HealthScape Advisors

This shift is being enabled by several factors. First, there is growing pressure to reduce administrative friction. At the same time, technology has evolved to support better understanding of provider-specific billing behaviors – including AI-driven claim selection and pattern recognition – and increasing ability to deliver real-time feedback.

But it also introduces new challenges:

• How to measure and credit “preventive savings”
• How to integrate PI into provider workflows
• How to align finance, actuarial, and operational models around upstream value

Key takeaway: Prepay is no longer the destination. It’s the bridge to something earlier, faster, and more collaborative.

6. Payment Integrity is at the center of a broader industry reckoning — and has a defining role in what comes next

Across sessions, there was a clear undercurrent: this isn’t just evolution — it’s inflection. The healthcare system is being forced to confront structural inefficiencies, and payment integrity sits directly in the middle of that tension.

From a practical standpoint, Shea Helmle highlighted how deeply fragmented the current system remains — particularly between Utilization Management (UM) and Payment Integrity (PI). The two functions operate with nearly identical logic but exist in silos, creating a “DMZ” where meaningful overpayment opportunity goes unaddressed. That gap is not theoretical — it’s operational, measurable, and largely unowned.

His message was pragmatic: the future doesn’t require entirely new capabilities — it requires connecting what already exists. Claims-auth matching, site-of-service validation, and enforcement of add-on codes represent immediate, tangible opportunities. More importantly, PI is uniquely positioned to lead this convergence, given its grounding in coding, payment logic, and data analysis.

“I think if you said, ‘What’s the perfect state?’ Get rid of prior auth. Use PI to enforce it. Plans would much rather pull the PI lever than the prior auth lever, because administratively, it ends up being a lot easier.”
— Shea Helmle, Buildr Health

But beyond operational fixes, another session — led by Lynn Garbee — zoomed out to something much bigger.

Her framing was blunt: healthcare today resembles the tobacco industry in the 1980s — widely understood to be inefficient and unsustainable, but too embedded to change organically. Administrative waste alone is estimated in the hundreds of billions, and much of it is tied up in the payer-provider “Cold War” — the endless cycle of denials, disputes, resubmissions, and negotiations that moves money without creating value.

There is a growing intolerance of the admin waste incurred by the payer-provider “cold war.”
— Lynn Garbee, Judi Health

There are two competing futures implied in her perspective:

1. One where PI continues to optimize within the existing model — getting better at detection, recovery, and negotiation, but ultimately reinforcing the same cycle
2. One where PI helps dismantle that cycle — moving upstream, embedding rules at the point of care, enabling real-time accuracy, and reducing the need for the entire denial-and-rework ecosystem

Key takeaway: Payment integrity, in its current form, is both part of that system and a potential lever to transform it.

Will payment integrity be a function that protects the status quo — or one that helps redesign it?

The disruption to the industry is already underway. Direct contracting, centers of excellence, and subscription-based care models are actively bypassing traditional claims flows. Employers are no longer passive — they are redesigning how care is purchased. And as that happens, the volume of claims — the raw material of traditional PI — is at risk of declining.

That raises a fundamental choice for the industry: Protect the status quo, or help redesign the future.

The answer likely hinges on how PI leaders respond to three realities:

1. The system’s inefficiencies are no longer hidden — they are quantified, visible, and increasingly unacceptable
2. The tools to change (AI, interoperability, real-time data) now exist — even if the operating models haven’t caught up
3. The pressure to act is coming from outside the traditional system — employers, regulators, and new market entrants

Payment integrity is no longer just about finding errors in the system. It’s about deciding what kind of system should exist going forward — and actively helping to build it.

What became clear over these conversations is that no single organization will solve these challenges alone. The path forward — from AI governance to pre-submission accuracy to payer-provider alignment — will be shaped by shared learning, honest dialogue, and collective progress.

That’s the real power of this community. And it’s exactly what POP Conference is evolving to support next: not just a moment to reflect, but a platform to build what comes next, together.