Agent group warns of tech reliability amid rollout
An Illinois State Farm agent trying out a new AI tool from the insurer found the technology gave her incorrect information — showing a wrong date for when she became a policyholder and incorrectly flagging a late payment, according to an article by The Wall Street Journal published Wednesday.
The Journal reported that the changes have met resistance from its sales force and customers after collecting responses from more than 1,000 readers. The 104-year-old mutual insurer is rolling out AI tools to its 19,000 agent offices under a “human + digital” strategy aimed at future-proofing its market dominance.
The changes include an AI tool that promises an “instant summary” of each customer’s household. A State Farm spokesman said agents can choose whether to use it, and 93% opted to do so. But agents and customers who responded to the Journal expressed widespread skepticism about the technology’s reliability.
“State Farm is a great insurance company,” one agent in Alabama told the Journal, speaking on condition of anonymity for fear of being fired. “We are a horrible tech company.”
Robert E. O’Connor Jr., general counsel at the National Association of State Farm Agents, which represents the sales force, said agents are “justifiably concerned about the dependability and reliability of State Farm’s technology.” He said rollouts over the past 10 years have had issues, including “a lot of outages.”
Agents also pointed to specific problems. A ChatGPT-style tool for agents has a “knowledge cutoff” of last year, according to screenshots reviewed by the Journal. Asked who won the Super Bowl, the most recent answer it confidently provides — the Kansas City Chiefs — dates to 2024.
Tom Duffy, a recently retired agent in Mesa, Ariz., said the insurer’s technology often creates more work for agents who have to sort out glitches. “It seems the more State Farm tries, the more work it creates for the agent’s office,” Duffy told the Journal. “AI will most likely not be any different.”
Customer responses were broadly negative. Many described chatbots and automated systems as “terrible,” “infuriating” and “it sucks,” the Journal reported. Joe Sonk of Moorestown, N.J., said the reason he has stuck with State Farm is the “great service by the local agent,” adding, “I don’t have to spend 10 minutes shouting ‘representative’ into the phone.”
Some customers said they would switch insurers if their agent’s office closed. A State Farm spokesman said the company does not aim to replace agents with AI but expects the reforms to result in larger agencies, suggesting smaller offices may consolidate or close.
The insurer has taken one step back from the changes. Following the internal uproar over contract modifications, State Farm extended a retirement-style benefit through 2031, subject to a “baseline performance expectation,” according to an email to agents reviewed by the Journal. The email, sent the day after the Journal’s article, said executives remained confident in the direction, adding: “The competitive landscape is changing. Standing still is not an option.”
State Farm is part of a broader industry push. Allstate CEO Tom Wilson told investors in April that his company is piloting direct sales done entirely by AI in three states and has given its agents an AI “customer engagement sidekick” that listens in and advises on calls.
Not all readers who responded to the Journal opposed the changes. Chris Tinen, a customer in St. Louis, Mo., said AI-driven cuts to insurance agents’ compensation were a painful necessity, comparing the shift to the disappearance of switchboard operators and elevator operators. “Commission sales reps have their days numbered,” he said.