On February 9, shares of the largest public insurance brokers saw a sharp decline – with stock prices for firms such as WTW, Aon and Arthur J. Gallagher all down on the day. Collectively, the U.S. public brokers, as measured by the MarshBerry Broker Composite Index, dropped 8.9% on February 9. The primary driver for this sector drop was linked to investor concerns that conversational AI is moving closer to the point of sale in the insurance distribution chain, with the potential to impact the traditional insurance distribution model – the broker.
MarshBerry doesn’t believe this is the moment when insurance brokers start to become obsolete. In fact, it may be the opposite. This may be the moment that brokers recognize the benefits that AI offers them to optimize and digitize their business processes, providing early adopters with clear competitive advantage.
AI enters the insurance buying journey
There were two headlines this week about AI’s most recent foray into the insurance distribution chain which set off an investor sell-off of public broker stock. The first was digital insurer Tuio’s launch of the first AI insurance app. The app can deliver personalized home insurance quotes, directly inside ChatGPT, without the need to interact with a broker. Purchasing capabilities will soon follow — meaning users will be able to complete the full insurance purchase process without leaving ChatGPT.
At the same time, Insurify, an online insurance comparison platform, announced what it described as the industry’s first ChatGPT insurance comparison app for auto insurance. The tool enables users to compare personalized price estimates and review insurers side by side within ChatGPT, drawing on Insurify’s data and customer-review base, before continuing with Insurify to complete the journey.
What sets these two developments apart from other “AI in insurance” headlines is the distribution mechanics. Through OpenAI’s App Directory (effectively an app store inside ChatGPT) third parties (like Tuio and Insurify) can embed real insurance products and workflows directly into the chat conversation – leading to a sale.
A balanced view of where AI disrupts and where it strengthens
It’s important not to overgeneralize what this might mean, because AI’s impact will definitely NOT be uniform. One must distinguish between areas where the buying journey is most compressible and where expertise remains central.
Similar to a flood – those at sea level will take the greatest hit. In this case, the clearest pressure point and those who will likely suffer the most from AI interaction is in standardized, price-led journeys (the commodity business, particularly in personal lines), where standard risks and large pools of underwriting data are readily available. But for the more sophisticated firms, this type of business represents a tiny percentage of total revenue, and AI is really just another step in the same direction that online comparison shopping and online buying brought on years ago.
The purchasing dynamic looks very different in commercial and specialty lines. Much of the profit pool for the public brokers is generated in areas where risk is nuanced and coverage is bespoke. Brokers that are closest to the risk will be better equipped to use AI to improve new client acquisition, reduce loss ratios and enhance retention. As the industry continues its race toward specialization, brokers are turning historical claims data from Iron Mountain into “data lakes” that produce proprietary underwriting insights. This will further drive the race from risk management to risk mitigation as brokers will be better equipped to build prescriptive solutions that reduce the frequency and severity of risk. Such solutions will be curated by industry verticals with unique and actionable risk-prevention insights and consulting capabilities, which will further drive the race toward the non-admitted market, expanding the focus on delegated underwriting authority, proprietary programs, and alternative markets, including syndicated risk.
At the end of the day, this is all just another step forward in the familiar direction of increasing efficiency and emphasizing expertise. More sophisticated brokers are already implementing AI to optimize and digitize their business processes — and the time saved is being redirected toward face-to-face and Zoom conversations with clients, enhancing relationships and service.
The net effect is that AI will not replace the fundamentals that M&A buyers have always valued in the insurance brokerage industry, such as leadership, culture, client relationships, and earnings quality. But it will increasingly influence how those fundamentals are achieved and how durable they’re perceived to be, which will in turn affect how buyers assess the type of growth they can have confidence in. From this perspective, AI stands to invigorate (not annihilate) the insurance brokerage industry.