Claims Adjusters Really Dislike AI. Here’s Why That Matters to You

Insurance claims adjusters are now some of the most skeptical workers when it comes to artificial intelligence. A recent Glassdoor analysis found that 98% of AI-related comments from adjusters on the site were negative. Across all jobs, that number was 53%.

Not every profession feels this way. Company leaders had the most positive comments about AI, with 67% showing up in the “pros” section of their reviews. Sales workers, product managers, recruiters, software architects, and data professionals were also fairly positive. Claims adjusters are at the other end of the spectrum, making up the most negative group Glassdoor tracked.

Why adjusters are worried about AI

Job postings for claims adjusters have dropped about 55% from their post-pandemic peak, compared to a roughly 36% decline across the rest of the labor market, according to Glassdoor’s analysis of Indeed data. Entry-level adjuster postings have fallen nearly 50% since early 2024, while the overall labor market fell 15%.

By 2026, entry-level postings for claims adjusters were at about half their 2025 level, Glassdoor reported.

Glassdoor researchers said this pattern matches what happens when experienced adjusters use AI to automate jobs that used to go to entry-level hires.

How insurers and adjusters are using AI

Insurers use AI to photograph and assess damage, sort claims, communicate with policyholders, and sometimes even approve and pay claims with little or no human involvement.

This isn’t always bad news for people filing claims. Anyone who’s waited days for an adjuster to inspect a damaged car or waited for a callback after a storm can see why automation is appealing. AI could potentially settle in seconds a claim that once took weeks to resolve.

Lemonade, one of the industry’s biggest AI adopters, says its AI system handles the initial notice of loss without human help 96% of the time and can automate about 55% of claims from start to finish. The insurer says its fastest claim took just three seconds.

State Farm announced this year that it’s testing an AI-powered virtual assistant for initial loss reporting, designed to simplify intake, triage, and documentation. The company calls this a “Human + Digital” approach.

Some public adjusters, the professionals who represent policyholders in disputes with insurers, are starting to use AI themselves. Companies now market AI tools specifically to public adjusters. These tools can analyze policies, compare an insurer’s estimate with the public adjuster’s, flag missing items, organize claim documents, and help prepare supplemental claims.

But some of the industry’s biggest AI supporters set limits on what they allow it to do. Lemonade says it doesn’t let AI automatically reject claims, and it admits that AI systems can produce inconsistent or biased results.

“I actually think AI makes experienced adjusters more valuable,” Lukas Esch, GTM leader at Strala, an AI-native insurance claims company, told Insurify. “When technology handles more of the repetitive administrative work, adjusters spend more time investigating complicated cases, communicating with policyholders, and making the decisions where their experience matters most.”

What happens when the AI gets it wrong?

Adjusters who now work alongside AI say they spend a lot of time fixing its mistakes. One Glassdoor reviewer said they spent “15–20 minutes per claim correcting AI-generated mistakes.” Another called the technology their employer used “bottom of the barrel.”

It’s not just errors. Adjusters are also worried that AI lacks the thoroughness needed for the job.

“An experienced professional would know when something deserves a closer look, even when every piece of the information given to them looks reasonable,” Alex Adekola, CEO and founder of Ready Adjuster, told Insurify. “This is where I think automation has to be used carefully because it doesn’t have the experience needed to know when to look closer. It only knows what to do based on the data it’s been given and doesn’t have the foresight to verify.”

Experienced adjusters investigate what happened, interpret policy coverage, assess damage, talk with repair shops and contractors, and know when a claim that looks simple actually isn’t. As insurers automate more routine work, the big question is whether technology can reliably spot the cases that still need human judgment.

“AI is very good at reviewing documents, organizing information, observing patterns, and helping simple claims move faster,” Esch said. But “insurance is full of edge cases, and an experienced adjuster may recognize something important because they’ve encountered a similar situation dozens of times over a 20-year career.”

The impact of getting a claim right or wrong goes beyond just the settlement amount. Claims can push homeowners and auto insurance premiums higher at policy renewal. Insurify data scientists predict the national average cost of homeowners insurance will reach $3,057 per year by the end of 2026 and the average cost of full-coverage car insurance will land at $2,242 per year.

What to do when an AI estimate looks wrong

For policyholders, the real question isn’t whether an insurer uses AI. It may soon be hard to find one that doesn’t. The real issue is what to do if the AI’s estimate looks wrong.

Policyholders should document the damage with their own photos and video. It’s also a good idea to keep repair estimates, contractor reports, receipts, and all correspondence with the insurer. Asking the insurer to explain how it calculated the estimate and whether a human adjuster can review the claim can help ensure the right result. If the insurer still isn’t handling it fairly, most state insurance departments have a complaint process consumers can use.

Related articles