Insurance Claims Adjusters and AI: Why the Backlash Matters for Builders
Published by AINave Editorial • Reviewed by Ramit
Insurance claims adjusters are the most vocal AI skeptics in the American workforce, with 98% of Glassdoor reviews mentioning AI being critical. For builders designing automated claims workflows, the lesson is clear: AI that misclassifies claims or hallucinates details doesn't just create rework, it erodes trust and drives away experienced staff.
The Glassdoor Backlash: 98% Critical Reviews
Glassdoor senior economist Chris Martin found that claims adjusters who mention AI in their reviews are critical 98% of the time, making them the biggest AI haters in the U.S. workforce. The complaints center on "AI-obsessed leaders forcing error-prone AI on them and their clients." This isn't abstract skepticism. It's grounded in daily experience with tools that add work instead of removing it.
Real-World Failures: Misclassification and Hallucinations
Ahmad Jackson, a former claims adjuster at a major insurer, saw this firsthand. His employer deployed AI for initial loss reporting, the process of setting up claims and gathering information after an incident. Instead of streamlining work, the AI produced misclassified claims that had to be rerouted and hallucinated details in summaries. When Jackson relayed those errors to claimants or their attorneys, he took the blame. He quit and moved to another carrier. "AI is getting things wrong," he told WIRED, "and it's implementing more work onto the adjusters."
These failures aren't edge cases. Sandy Avina, a claims adjuster turned consultant, notes that a smudge in a document can trigger a hallucination, leading to an incorrect payout. Customers don't realize AI caused the error, so they blame the adjuster. That erodes trust in both the tool and the person using it.
The Labor Market Signal: Fewer Adjusters, More Automation
The Bureau of Labor Statistics projects 18,900 fewer claims adjusters by 2030, with a 21% drop between May 2025 and May 2026. Entry-level postings have fallen 50% since 2025, according to Glassdoor. The BLS cites technology as a major driver. Meanwhile, Lemonade's AI Jim chatbot handled 96% of initial reports by year-end, with automation handling roughly 55% of all claims. State Farm takes a more cautious approach, emphasizing a mix of human and digital expertise.
For builders, this creates a tension. The market is pushing automation hard, but the frontline workers who have to use these tools are deeply skeptical. If your AI product adds friction or errors, you'll face resistance that undermines adoption.
What Builders Should Do Differently
The claims adjuster backlash offers concrete lessons for any team building AI for regulated, high-stakes workflows.
First, design for augmentation, not replacement. Geoffrey Conrad, a claims executive, puts it bluntly: "AI is just a tool. It should never be given the keys." Builders should ensure that AI handles routine administrative tasks like extending rental car bookings, while complex decisions remain with humans.
Second, invest in guardrails against misclassification and hallucination. Claims summaries and medical record analysis are high-risk areas. A single hallucinated detail can lead to an incorrect payout and a damaged relationship. Implement human-in-the-loop review for any output that affects policyholders.
Third, prioritize explainability and auditability. When an AI makes a mistake, the adjuster needs to understand why and be able to correct it. If the system is a black box, trust erodes quickly.
Fourth, pace deployment. The Glassdoor data shows that when workers sense layoffs looming, their reviews become increasingly anti-AI. Rolling out automation too fast without addressing job security concerns will poison adoption.
Caveats and Limitations
This analysis is based on a single Wired feature and its cited statistics. Real-world outcomes vary by insurer, product line, and region. BLS projections are historical and subject to revision. The pace of AI adoption can change with regulation and technology maturation. Claims outcomes depend heavily on data quality, model correctness, and integration with human workflows. The benefits of automation are contingent on robust governance and thoughtful deployment.
