Over the last few months, I’ve been learning about Workers’ Compensation.
Most people (rightly) expect AI to transform healthcare, from early risk detection to treatment optimization. When they imagine this transformation, they usually think of hospitals and health plans. Yet, in the adjacent world of P&C insurance, Workers’ Compensation also operates as a form of health insurance for employees injured in work-related settings. It often builds on a managed network model where a risk-bearing insurer or self-insured employer finances care from premiums, directs injured workers into approved networks, and pays providers largely under fee schedules. And, just as in traditional healthcare, there is wide variation in provider performance, guideline drift in utilization, constrained analytics due to documentation limitations, and incentives that fixate on unit price rather than overall outcomes. Care is generally delivered within the same healthcare settings used for non-work injuries, and employer costs for Workers’ Compensation are expected to be roughly $100 billion in 2025.
Given these parallels, can Workers’ Compensation carriers, TPAs, and their consultants leverage and benefit from AI capabilities originating in the classical healthcare ecosystem?
Yes. As we established, Workers’ Compensation is effectively a managed health network. Moreover, medical costs and wage replacement costs are tightly coupled: poor clinical management and slow return to work increase both medical outlay and the wage replacement expense. Conversely, AI-enabled improvements in clinical management positively impact both spend buckets; AI may be the most powerful lever to hit Workers’ Compensation in decades.
Tactically, AI impacts clinical management through three functions. First, in intake and triage: deploying AI risk stratification to prioritize complex cases and trigger early nurse involvement. Second, in medical management: aligning authorization workflows to guideline-concordant pathways, steering to proven network providers, and integrating pharmacy and bill review rules directly into adjuster tools. Third, in analytics and governance: maintaining a longitudinal clinical and financial record, which supports continuous learning and transparent reporting.
In exploring this space, I’ve come to believe that the biggest untapped opportunity in Workers’ Compensation is not marginal improvements in business processes but using AI to turn carriers and TPAs into data-driven clinical managers of medical care.
