AI adoption surges in P&C operations, but training budgets lag

The biggest risk in P&C operations is no longer whether insurers will use AI. It is whether they will train enough people to use it safely, consistently and at scale. Covenir’s second annual Insurance Operations Leaders Trends Report found that 70% of organizations already had AI running in live operations, yet 20% were cutting training budgets and only 7% were protecting them.

The survey, fielded in February and March 2026 among 152 U.S.-based insurance operations decision-makers at carriers, MGAs and insurtechs, shows a market that is spending on tools faster than it is spending on enablement. Nearly 68% of organizations saw budget increases in 2026, but that did not translate into relief on the floor. Sixty-five percent of operations leaders said their teams were more stretched than they had ever been, while 91% of executives said the same. The strain was sharpest at mid-sized carriers with 500 to 999 employees, where 83% reported heavy pressure.

That creates a direct business risk for AI rollouts. Training cuts can slow adoption, because staff who are not coached on workflows, exception handling and model limits are less likely to trust the system or use it correctly. They also weaken controls, since frontline teams are more likely to miss errors, over-rely on automated suggestions or bypass new processes entirely. When 29% of respondents expect outsourcing partners to have AI embedded in their services, the danger is even clearer: carriers may end up renting intelligence from vendors instead of building internal capability around it.

The operational pain points are concentrated in familiar places. Forty-two percent of respondents said brand promise breakdowns most often happen at first notice of loss, the moment when claims intake, communication tools and manual triage face the most pressure. In that same environment, 81% said operational insights are critical to business success, but 47% either were not using their data or could not translate it into decisions. That gap matters because it points to poorer outputs, slower decision-making and a heavier dependence on vendors to interpret what the carrier’s own systems should already be telling it.

Covenir CEO David Squibb said the results showed a widening gap between technology investment and operational readiness. In a June 12 blog post, he added that 30% of leaders at organizations with 1 to 199 employees were not confident their operations would consistently deliver on the experience their brand promised. He also cited J.D. Power data showing that 80% of auto insurance customers who have a poor claims experience switch or plan to switch carriers. For P&C insurers, that makes training budget decisions a claims issue, a retention issue and an AI strategy issue all at once.