Artificial intelligence is rapidly becoming part of the healthcare industry's risk adjustment strategy. From analyzing clinical documentation to identifying potential documentation improvement opportunities and streamlining chart review, AI is helping payers and providers extract more value from data than ever before. In fact, a 2025 survey conducted by the National Association of Insurance Commissioners (NAIC) found that 84% of health insurers reported using AI or machine learning technologies.
Organizations are applying AI and NLP across a range of functions that support risk adjustment and payment accuracy, including utilization management practices, implementing disease management programs, prior authorization processes, fraud claim detection, and medical provider fraud detection.
Yet as adoption accelerates, one reality is becoming increasingly clear: success depends not on replacing human expertise, but on enhancing it. Healthcare organizations must determine how to responsibly integrate AI into existing workflows while maintaining confidence in the quality, accuracy, and integrity of their results.
Unlocking value from healthcare's most complex data
Risk adjustment programs rely on vast amounts of information generated across the healthcare ecosystem. Much of that information exists in unstructured formats—including physician notes, discharge summaries, and medical records—which can be difficult and time-consuming to review manually.
Natural language processing (NLP) and AI play complementary roles in helping organizations unlock value from this data. NLP excels at extracting clinically relevant information from unstructured documentation, while AI can identify patterns, prioritize opportunities, and surface insights across large volumes of information. Together, these technologies help organizations uncover coding opportunities, identify documentation gaps, and focus resources on the records most likely to drive value.
This distinction is important. Some opportunities, particularly those found deep within clinical documentation, require the precision of mature, well-trained NLP to identify relevant evidence and supporting context. AI can then help organize, prioritize, and accelerate the review process. Rather than replace competent personnel, the human in the loop becomes more efficient and effective. The result is a more efficient approach that allows clinicians and coders to spend less time searching through records and more time evaluating meaningful findings.
Integrating AI into risk adjustment workflows
The question is no longer whether AI belongs in risk adjustment. The question is how organizations can use it responsibly at scale.
As AI becomes more embedded in coding and documentation workflows, success will depend on trust as much as technology. Organizations need visibility into how recommendations are generated, confidence in the supporting clinical evidence, and governance processes that ensure AI-driven insights align with compliance and business objectives.
The most successful organizations will build the right balance between automation, transparency, and expert oversight.
What this means for health plans
For payers, AI can help improve productivity, streamline chart review workflows, and uncover clinical insights that may otherwise be overlooked. When supported by reliable governance frameworks, AI can enhance performance while helping plans maintain confidence in risk adjustment outcomes.
Successful payer organizations are focusing on:
- High-quality, clinically validated data inputs
- Explainable AI that surfaces supporting clinical evidence alongside coding recommendations
- Ongoing monitoring and model refinement
- Strong security, privacy, and compliance foundations
These principles help ensure AI delivers measurable value while supporting regulatory requirements and organizational objectives.
What this means for providers
Providers are increasingly experiencing the benefits of AI-driven risk adjustment as well.
AI and NLP can support prospective risk adjustment efforts by helping care teams identify existing and potential conditions, prioritize what should be evaluated during a patient visit, and support more effective documentation processes. By surfacing relevant clinical insights before an appointment, these technologies can help providers focus on the most meaningful opportunities while reducing administrative burden. Beyond that, effective documentation enables greater insight into the complexity of disease, leading to opportunities to improve care management.
The key is ensuring that technology supports rather than disrupts clinical workflows. Providers need solutions that deliver actionable insights while preserving physician autonomy and clinical decision-making.
Looking ahead
As AI adoption accelerates across healthcare, the organizations that realize the greatest value will be those that pair sophisticated technology with experienced human expertise.
The future of risk adjustment isn't human versus AI. It's human and AI working together—leveraging technology advances to transform complex data into actionable insight while keeping clinical judgment at the center of every decision.
About the author
Dr. Summerpal Kahlon leads the product management group for Cotiviti Health Enablement, working across products and functions to guide clinical strategies and policies. His career spans over 20 years of experience in diverse health care settings, businesses, and markets. Dr. Kahlon has deep expertise in value-based care, quality, and risk adjustment, and brings unique insight to help customers make the most comprehensive, informed, and clinically relevant decisions for their populations.