NCCI's 4% Severity Spike Tests AI Claims Speed in Workers Comp

Workers comp medical and indemnity severity each rose 4% in 2025 even as AI claims tools pushed a growing share of losses toward early closure. That pairing compresses the first 12 to 24 months of a loss triangle while underlying severity accelerates.

Development factors calibrated on that shortened window will understate ultimate losses on the complex claims AI closes last, and the compression looks like improvement in every headline measure.

Key Takeaways

• 4% growth in both medical and indemnity severity in 2025, against a lost-time frequency decline of only 2%, milder than the 2% to 3% long-term average.
• A 91 calendar year combined ratio against a 102 accident year figure. The composite is held down by favorable development released from older accident years, not by current-book performance.
• 12% to 19% lower total claims costs and 35% to 50% faster cycle times reported by carriers deploying advanced AI in workers comp claims operations, concentrated in low-to-medium severity claims.
• 20% to 45% higher total incurred on claims first reported more than seven days after injury versus within three days. Compressing reporting lag mechanically shifts claims into the fast-closing bucket.
• 59% of eventual $1 million claims now reach that threshold within 24 months, up from 27% in accident year 2003, so the tail was already emerging faster before closure speed changed.

What the 2025 Numbers Say

NCCI delivered the 2025 data at its Annual Insights Symposium in Orlando on May 12, 2026, alongside a calendar year combined ratio of 91, the twelfth consecutive year of underwriting profitability for private carriers. Chief Actuary Donna Glenn put the composite in context: "There's not a single number that defines the workers compensation system. Behind this year's combined ratio of 91, factors such as industry mix, state differences, and carrier variation are all shaping results."

The underlying pattern breaks the line's usual shape. Lost-time claim frequency fell 2% in 2025, a milder decline than the 2% to 3% long-term annual average of the past two decades, while both components that are supposed to offset frequency improvement moved the other way: medical severity rose 4% and indemnity severity rose 4%.

Calendar year and accident year tell different stories. NCCI's accident year 2025 combined ratio came in at 102, above breakeven, with the calendar year 91 held down by favorable development released from older accident years. Net written premium fell 0.2%, and the industry's aggregate reserve position remains $14 billion redundant.

Metric	2025 value	Direction
Lost-time claim frequency	-2%	Improved, milder than long-term trend
Medical claim severity	+4%	Accelerating
Indemnity claim severity	+4%	Accelerating
Calendar year 2025 combined ratio	91	Profitable, 12th consecutive year
Accident year 2025 combined ratio	102	Above breakeven
Net written premium change	-0.2%	Contracting
Industry reserve position	$14B redundant	Cushion from older accident years

A carrier or regulator reading only the calendar year figure could reasonably conclude the line is stable. The accident year result, paired with 4% severity growth on both components, says the current book is running hot and the redundancy cushion from older years is what keeps the composite comfortable.

Two Populations, One Set of Factors

AI claims platforms are compressing the front end of development at the same time, and the two effects combine into a triangle that reads as improvement.

CCC Intelligent Solutions completed its $730 million acquisition of EvolutionIQ in January 2025 after announcing it in December 2024, and by early 2026 the platform guided claims at nine of the top 15 US disability carriers, with expansion underway into workers compensation. Deloitte's 2026 survey found carriers deploying advanced AI in workers comp claims operations reported 12% to 19% lower total claims costs and 35% to 50% faster cycle times.

The mechanism is specific. AI triage accelerates first notice of loss, routes claims within hours instead of days, and drives earlier medical intervention. NCCI has documented that claims first reported more than seven days after injury cost 20% to 45% more in total incurred than claims reported within three days, so a platform that compresses reporting lag mechanically pulls a share of the book into the low-cost, fast-closing category.

That is where the effect stops. Complex musculoskeletal claims requiring multiple surgeries, occupational disease claims with delayed manifestation, and claims complicated by comorbidities in an aging workforce continue to develop in months 12 through 60, a window sitting almost entirely outside the AI-optimized closure period. That segment is also where most of the 4% medical severity growth is generated.

Put the two together and the triangle bends toward what looks like genuine improvement. Early development factors shrink because a larger share of claims closes inside the accelerated window. Later factors, applied to a shrinking population of complex claims the platform never touched, do not shrink at the same rate, and in a rising-severity environment they should be widening. An actuary blending the last five years of experience, which is how NCCI's methodology and most carrier selections work, pulls the recent AI-affected periods through the whole triangle, including the tail maturities where the effect never operated.

The signature to test for is narrow: a first-diagonal compression on the order of 8% to 12% against a carrier's own five-year average, paired with 24-to-60-month factors that hold steady or widen rather than following the first diagonal down. That combination is not one improving process. It is two populations blended into a single set of factors representing neither.

The tail was already moving before AI arrived. NCCI's fast-emerging large claim data shows the share of claims eventually exceeding $1 million that reach the threshold within 24 months grew from 27% in accident year 2003 to 59% in accident year 2023.

ULAE Inherits the Distortion Twice

Unallocated loss adjustment expense reserving carries the same bias and gets less attention, because ULAE is typically built from a historical ratio to paid losses or open claim counts rather than modeled on its own.

AI-accelerated closure front-loads handling activity and expense recognition into a claim's first months, which is when the platform does most of its work: intake triage, routing, automated documentation. The claims still open at 12, 24 and 36 months are disproportionately the complex cases the platform did not resolve, the ones requiring nurse case management, independent medical examinations, vocational rehabilitation and litigation support. Those are ULAE-intensive activities concentrated in exactly the tail AI does not touch.

A ratio calibrated on a pre-AI period understates that population twice over. It reflects a claims mix that included more low-complexity claims still open at later maturities, diluting average handling cost per open claim. And as AI closes the easy claims earlier, the open inventory at any evaluation date concentrates further in above-average-cost cases, so the same ratio applied to a smaller but costlier population produces a larger understatement with each successive accident year.

The distortion does not stop at a single carrier's balance sheet. State rate filings and actuarial opinions relying on industry-average development factors, whether NCCI's advisory loss costs or a bureau state's promulgated factors, inherit whatever blend of AI-adopting and non-adopting carriers sits inside that countrywide data.

NCCI extends loss costs, rates and expected loss rates to three decimal places for every classification code in the 2026 cycle, but that precision operates on the same blended development data. A filing showing an unusually favorable loss trend alongside accelerating industry severity is describing two things at once, and the countrywide factors will lag the divergence the same way an individual carrier's five-year blend does.