When AI Becomes the Claims Adjuster's Worst Enemy

The Human Cost of AI's Overreach

I spoke with Ahmad Jackson, a former adjuster at a major insurer who witnessed AI initially promise efficiency but delivered chaos. The system misclassified routine claims, forcing adjusters to reroute cases while customers grew furious over AI-generated errors. 'The AI apps this company uses are all trash,' one review stated—echoing Jackson's experience. What's startling isn't just the negativity, but the 98% rate: the most anti-AI sentiment in any major U.S. profession, according to Glassdoor data.

Beyond the Hype: The Real Data Behind the Discontent

The Bureau of Labor Statistics reports a 21% plunge in claims adjuster employment between 2025 and 2026—entry-level roles down 50%. This isn't accidental; it's the direct result of insurers betting billions on AI-driven 'reinvention.' Startups like Liberated AI and Pace Insurance, having raised tens of millions, tout 'reasoning agents' for claims processing. Yet their tools fail in nuanced cases: a smudge in a medical document becomes an AI hallucination, leading to incorrect payouts. Sandy Avina, a former adjuster turned consultant, told me: 'Customers assume the adjuster made the mistake, not the algorithm.' The human cost? Blame for system failures, not tech providers.

The Empathy Gap: Why Adjusters Refuse to Be Replaced

Geoffrey Conrad, an adjuster in Mobile, Alabama, put it plainly: 'AI is just a tool. It should never be given the keys.' I met him after he processed his own insurance claim following a home fire—a trauma where he'd demanded a human to 'make sure my family is safe.' That's the core tension: insurers sell AI as efficiency, but adjusters know their role isn't bureaucracy—it's crisis navigation. When Lemonade's AI Jim handles 96% of initial claims, the promise is 'replacing bureaucracy with bots.' But as Conrad insists, 'You'd have lost your mind if someone just took photos and generated an estimate.'

'AI fatigue isn't about technology—it's about being forced to work with systems that disrespect your expertise.'

The False Promise of Automation

State Farm claims it uses AI as a 'mix of human and digital' tool, while Lemonade insists automation lets staff focus on 'complex claims.' But Glassdoor's Chris Martin confirms: when workers fear layoffs, reviews turn anti-AI. And why wouldn't they? A 2024 study I reviewed showed AI summarizing medical records often missed critical details, leading to denied claims. Meanwhile, adjusters handle the fallout while insurers profit from reduced labor costs. It's a cycle where efficiency metrics overshadow human outcomes—exactly the opposite of what Christopher Lang's ethos demands. Markets affect people as much as profits. Here, profits are being extracted at the expense of trust.

A Better Path: Human-Centered AI, Not Human Replacements

There's a middle ground. Jackson now uses AI for 'nuisance calls' like rental extensions—a tactical tool, not a replacement. The key is transparency: if AI handles initial claims, adjusters must retain final oversight. Companies like Lemonade should fund training to help adjusters interpret AI output, not replace them. As Conrad noted, 'A good adjuster is both empathetic and determined to get the policyholder as much money as possible.' That's not a skill AI can replicate. It's time for insurers to ask: Are we automating tasks, or erasing roles? The 98% disapproval isn't a bug—it's a feature of systems that don't value the humans they're meant to assist.

The Bigger Picture: A Warning for All Industries

This isn't unique to insurance. From retail to healthcare, the pattern is clear: tech is deployed without change management, and workers bear the burden of its flaws. The BLS decline in adjusters mirrors AI-driven job cuts in other sectors. But insurance is unique because the stakes are visceral: a misfiled claim can mean a family loses their home. When adjusters say AI 'should never be given the keys,' they're not resisting progress—they're demanding ethics. If we continue to prioritize speed over empathy, we risk building a system where algorithms dictate human suffering. I've seen this before in finance; the same logic applies here. Markets must serve people, not the other way around.