AI Disintermediation

On Feb. 9 public insurance broker stocks fell by 9% on the news that two digital companies had launched chat-bot assistants using ChatGPT’s tools. While the valuation of those brokers’ stocks have largely recovered, at least a few analysts suggest that thinking may be premature.

In a report by BofA Global Research, “Putting numbers around insurance agent/broker AI disintermediation risk,” analysts reported that at least $15 billion of independent agency commissions and broker fees-those considered “low complexity”-are at risk for some disintermediation. Analysts estimate in total U.S. independent agency commissions/broker fees were in excess of $100 billion in 2025.

The report looked at six carriers catering to small business and personal lines, including Travelers (Select Accounts and Personal only), Hartford (Small Business and Personal only), Progressive, Cincinnati Fin’l, Hanover and Selective. In review of those carriers, BofA said they believe that over $15 billion of commissions, largely skew toward low-complexity, are at-risk.

“There are clearly other carriers-most notably Liberty Mutual-who generate additional billions in commissions that also can skew toward low complexity. Public insurance agent/broker disclosure tends to be very thin, making it difficult for investors to estimate what proportion of their commissions skew toward low complexity,” the report said.

The report’s authors added that they are not suggesting a “mass disintermediation of the entirety of a business” but believe a “non-immaterial minority of the business” could be a risk.

“Our view is that large-language model digital agents can effectively do a non-immaterial portion of the work currently provided by 20-30k independent agents across the United States,” the report stated. But nothing is certain. “As with many theses regarding technological innovation, the future is hard to predict,” the authors wrote.

For now, it seems large commercial risks are not “at-risk” for AI disintermediation. The complexity of such risks and the value that an agent and broker provide to these insureds cannot be replaced.

Still, the report noted there could be a “deflation” in large-case business.

“While we do not expect large insurance buyers to use digital agents to purchase insurance, we do expect A.I. will de-mystify insurance markets,” the analysts wrote. “Sophisticated insurance buyers will benefit from the deflationary impact of technological innovation, potentially reducing pricing power for their services.”

Topics InsurTech Data Driven Artificial Intelligence