Every day, payers deploy artificial intelligence (AI) into claims and prior authorization with the hope of eliminating delays, rework, and provider friction. Yet, too many are still seeing clean claims sent back for missing context, prior authorizations stalled over data that already exists, and a need for continued manual reviews because systems can’t agree. These disappointments stem from one root cause: payers are deploying AI as tools, not as infrastructure.
Without integration across data, decisions, and workflows, organizations are failing to realize the promise that AI offers to lower the cost of administration and improve the quality of their relationships with both providers and members.
Here’s the solution: payers need to directly embed AI into the ways work gets done, and decisions are made. That means embedding AI as the connective layer linking data, decisions, and workflows across five key pathways:
- workflow orchestration,
- trust and governance,
- data modernization,
- value realization, and
- operating model redesign – particularly how decision rights, accountability, and human oversight are defined in AI-enabled processes.
AI as the orchestrator inside workflows
The age of proof-of-concept pilots is over – we have ample proof that AI can work. Now is the time to focus on deploying AI technology at scale throughout the organization. Inserting AI inside workflows embeds the technology directly into the decision-making process, offering the potential to eliminate costs from delays or rework, connect steps that are typically disjointed, and coordinate decisions across systems that historically operate in silos.
For example, during claims intake, AI can interpret clinical documentation, flag missing elements before submission, and route clean claims directly to adjudication. In prior authorization, it can align clinical criteria, patient history, and policy rules in real time, reducing back-and-forth with providers.
Governance breaks in fragmented AI environments
In payer organizations, many of the highest-stakes decisions are already under scrutiny from regulators, providers, and members. When AI is deployed across disconnected tools, decision logic varies by workflow, auditability becomes inconsistent, and organizations struggle to explain or defend outcomes.
Without a consistent approach to oversight, payers are forced to manage risk at the use-case level, slowing adoption and limiting impact.
Treating AI as infrastructure changes this dynamic. Governance becomes embedded into the system itself, with consistent standards for explainability, bias monitoring, and auditability applied across workflows. Instead of managing isolated risks, organizations can enforce decision consistency at scale – not only strengthening compliance, but also reducing provider friction and accelerating adoption and building trust in how decisions are made.
Modernizing data around access, not perfection
Legacy data environments remain a barrier for many payers, particularly where clinical, claims, and member data sit in separate systems. A common assumption is that large-scale data transformation must happen before AI can be deployed.
In practice, progress comes from improving interoperability between legacy systems and AI. Payers can enable AI to access and connect data across systems without waiting for a full rebuild. This is especially important for use cases like care management, where insights depend on linking claims history, clinical data, and social determinants.
By focusing on making data usable and accessible, organizations can begin operationalizing AI while continuing to modernize over time rather than delaying impact in pursuit of a perfect future-state architecture.
Making value measurable
AI investments in payer organizations often stall because value is not clearly defined or tied to business outcomes. Programs are treated as innovation efforts rather than performance drivers.
A more effective approach is to anchor AI initiatives to payer-specific metrics from the outset. These include reductions in claims rework rates, prior authorization turnaround time, call center volume, and administrative cost per member.
Leading organizations define these desired outcomes upfront, track performance continuously, and tie accountability to functional leaders across claims, clinical operations, and member services. When value measurement is embedded into the operating model, leaders gain the confidence to scale beyond pilots and treat AI as a core lever of operational performance, not a discretionary technology investment.
The competitive imperative for payers
Healthcare administration is at a turning point. In an AI world, the differentiator will not be which payers adopt the technology, but which integrate it most effectively across their operations.
Organizations that continue to deploy AI as disconnected tools will see incremental gains. Those that redesign workflows around connected, intelligent systems can reduce administrative burden, improve provider experience, and respond faster to member needs while maintaining consistent governance and trust at scale.
In a market defined by tight margins, regulatory scrutiny, and rising consumer expectations, AI’s role is not just to automate tasks. It is to become the foundation for how decisions are made and executed across the payer enterprise.
J.D. Brewer
As a leader in KPMG’s Health and Government Solutions team, J.D. Brewer assists clients with transformation projects related to technology and operating models. His experience includes assessment and implementation of new operating models for the front, middle, and back office of payers and integrated delivery networks. His focus on sustainability of change has helped organizations obtain topline growth while reducing the cost of sale and delivery.
Saurabh Goyal
Saurabh Goyal is an operations transformation leader focused on strategy, growth, and improvement for Health insurers and Pharmacy organizations. He specializes in business model innovation, operating model transformation, value-based care, digital health, and data & analytics. He helps clients design, implement, and communicate the next-generation member-centric care models that are digitally enabled, focusing on member back experience, reducing the total cost of care, and improving efficiencies while establishing new and practical metrics to drive business outcomes.
