According to WIRED, one group on the job-search and workplace-review platform Glassdoor has a far stronger aversion to artificial intelligence than any other group.
“Forcing AI into everything so that people are no longer expected to use their own thoughts and brains is truly repulsive,” one representative review said. Another stated: “Stop forcing AI on everyone.” Someone else put it more directly: “All of the AI applications this company uses are garbage.”
Surprisingly, these reviews were not left by journalists or teachers, but by insurance claims adjusters.
A recent Glassdoor study found that as many as 98% of insurance claims adjusters who mentioned AI in their reviews were critical of it, making them one of the most AI-averse professional groups in the U.S. labor market.
According to the study, claims adjusters most often complained that “AI-obsessed management forces error-prone AI on employees and customers.” When AI causes problems in actual work, claims adjusters are often still the ones who have to clean up the mess.
Ahmad Jackson experienced this firsthand.
About a year ago, he worked in the claims department of a large insurance company. The company had decided to introduce AI into the first notice of loss process, using AI to create a claim and collect relevant information when a customer first reported an accident.
The company had hoped the technology would benefit both employees and policyholders: simple claims could be handled more quickly, while complex cases would be passed to actual staff.
But that was not what happened.
Jackson said he and other adjusters were suddenly faced with a large number of cases that had been misclassified by AI, forcing them to reroute the cases to the correct departments.
He also repeatedly found so-called “hallucinations” when reading AI-generated claim summaries. When he inadvertently repeated this incorrect information to claimants or their lawyers, he was the one who faced the customers’ anger.
Before long, Jackson resigned and moved to another insurance company.
“AI gets things wrong a lot,” Jackson told WIRED, “and it actually gives claims adjusters more work.”
Geoffrey Conrad, a claims-industry executive from Mobile, Alabama, also said that clear signs of “AI fatigue” had emerged within the industry. “We’ve had enough of all this AI being shoved into our work,” he said.
Glassdoor senior economist Chris Martin had not expected insurance claims adjusters to be so hostile to AI. When he saw the results, he said he even rechecked the data before beginning a deeper investigation.
Martin ultimately concluded that the insurance claims industry is undergoing a major adjustment.
In 2024, the U.S. Bureau of Labor Statistics (BLS) projected that the number of insurance claims adjusters in the United States would decline by 18,900 over the next decade, a decrease of about 5%.
According to BLS data, employment in the industry then fell sharply by 21% from May 2025 to May 2026.
The situation is even more severe for adjusters at the beginning of their careers. Glassdoor data shows that the number of entry-level job postings in the insurance claims industry has fallen by 50% since 2025.
The BLS identifies technological development as one of the major factors driving the decline in employment.
Over the past several years, numerous AI startups have entered the insurance industry. Companies such as Liberate and Pace have raised millions of dollars in funding with visions of “reinventing” insurance, while traditional insurers have continued expanding their use of AI, allowing artificial intelligence to handle an increasing number of claims.
In practice, this could mean that when customers submit claims, they no longer communicate with a human but instead interact directly with a chatbot.
For cases involving property damage, insurers may stop sending employees to inspect the site and instead have AI analyze photos or videos provided by customers to estimate the payout.
Hundreds of pages of medical records can also be given to AI for summarization and key information extraction.
Insurance companies’ AI systems can even process photos, videos, supporting documents, receipts, and other materials uploaded by policyholders, then complete the review and issue payment within seconds.
That is exactly what insurance company Lemonade hopes to achieve.
Since its founding in 2015, Lemonade has repeatedly said it wants to use “bots and machine learning” to replace the cumbersome administrative processes of traditional insurance.
By the end of last year, the company’s proprietary chatbot AI Jim had handled 96% of first notices of loss, while automated systems had participated in processing about 55% of all claims.
By contrast, traditional insurers such as State Farm emphasize that claims work still requires a combination of human experience and digital technology.
Some claims adjusters nevertheless worry that they are effectively helping train the AI that will replace them in the future.
Lemonade spokesperson Paul Staats said: “AI will affect many jobs and pose a threat to many existing professions in the insurance industry and throughout the economy.”
At the same time, he believes automation can allow employees to focus their “empathy, care, and expertise” on the most complex claims.
Staats also said that Glassdoor’s report “should be taken seriously by the entire insurance industry.”
“There are no jobs right now, and there isn’t much of a positive employment outlook,” said Glassdoor senior economist Martin.
According to Glassdoor data, when employees feel that layoffs are approaching, the reviews they post tend to become increasingly opposed to AI.
Martin said employees also express strong skepticism and dissatisfaction with AI when they believe their company is forcing a poor-quality AI product on them or their customers.
Claims-industry executive Conrad said that many of his colleagues genuinely feel they are being replaced by AI models.
“AI is just a tool,” he said. “It should never be given control.”
Of course, not every claims adjuster believes AI is without value. Jackson said AI can genuinely help with some administrative tasks. For example, he uses AI for telephone matters he considers relatively “trivial,” such as extending a customer’s rental-car reservation by a few days.
State Farm spokesperson Justin Tomczak also said the company wants “to provide State Farm agents and employees with better tools so they can spend more time doing what matters most: helping customers.”
However, Sandy Avina, who formerly worked as a claims adjuster and is now an insurance-industry consultant, said claims adjusters have “little confidence” in AI’s output. For example, if a document submitted by a lawyer contains a blemish or a blurry mark, it could cause AI to hallucinate and ultimately result in an incorrect payment.
Similarly, if AI omits a key piece of information while summarizing a medical report, it could lead to an incorrect compensation amount.
More troublingly, customers usually do not know that AI is the true source of confusing or incorrect information. They often assume that the claims adjuster handling the case made the mistake.
Conrad knows that many people do not have a favorable impression of insurance claims adjusters, and may even see them as “cold vultures” circling after disaster strikes.
But he insists that a good claims adjuster should be empathetic and willing to do everything possible to help policyholders receive the compensation they deserve.