Rapid MGA market growth is outpacing the processes meant to keep insurance data usable across the entire delegated authority value chain. With such market expansion comes more data, more processes required and more passing of data from one stakeholder to another, said Tony Russell (pictured), chief revenue officer at VIPR.
The ideal scenario for MGAs, brokers, carriers and reinsurers to process risk, premium and claims data under delegated agreements would be the same flow of information, said Russell. As larger MGAs and brokers automate their data process, this can help strengthen their carrier partnerships.
The MGA market has grown at more than 20% year-on-year, and has attracted regulatory attention to ensure governance processes can adapt.
By the time a bordereau has passed through four or five systems on its way to a reinsurer, Russell said, the information can be “almost indecipherable”.
MGA onboarding and accreditation, checks against contractual limits and territories, and sanctions screening have historically been treated as “a side-of-the-desk job”, Russell said, even as delegated authority becomes a primary route into specialty lines for more managing agents.
Sanctions screening is particularly difficult to perform manually at current volumes. Russell cited a one-million-line bordereau processed by VIPR as an example of the scale involved, arguing that no human team could check every line within a single reporting month.
Large UK brokers are building the capability to process bordereaux and pass the resulting data to carriers affordably at scale. That type of investment is in preparation for future growth through acquisitions and delegated portfolio increases, said Russell.
The Financial Conduct Authority’s (FCA) expanded review of MGA and coverholder governance, with findings expected in early 2027, sits behind much of the current focus. Russell expects the eventual model to resemble SOC 2 Type 2 compliance, with annual independent audits against defined controls. Any broker taking on bordereaux processing itself would therefore need to plan for an ongoing compliance cost rather than a one-off regulatory exercise.
Any firm building or buying delegated data capability also inherits questions about how artificial intelligence is being used across the technology supply chain.
Russell said firms are not examining vendors’ AI practices closely enough. He cited an industry survey suggesting that 70% of software buyers had not assessed how their technology vendors were applying AI to the tech.
The questions extend beyond whether a system uses AI. Firms also need to understand whether it can be disabled, how customer data is separated and where human oversight remains necessary. That scrutiny becomes particularly important where AI could influence claims decisions.
“If you’re asking a chatbot to assess a claim and you don’t have a human in the loop on that assessment, then I think you may have a difficult outcome to explain,” he said.
The legal and governance implications of AI for insurers remain unsettled. Russell expects responsibility to become clearer only after high-profile failure tests where liability sits between an insurer and its technology supplier.
Russell’s broader argument is that the technology needed to improve delegated data has existed for years, but the market has lacked the willingness to overhaul established processes. He said that resistance is tied to a wider conservatism within an industry built around managing risk.
Private equity ownership of MGAs is starting to change that dynamic, he said, by putting greater pressure on firms to operate efficiently rather than through multiple layers of committees.
Collecting and cleaning data is not, on its own, a competitive advantage. The distinction lies in whether firms can use that information to improve underwriting, oversight and decision-making. Removing the 90-day lag of data transparency is now supporting every stakeholder in the chain.
The FCA’s regulatory priorities for delegated authority and claims are moving in the same direction, with growing emphasis on firms evidencing how their oversight and data translate into decisions. Cost is only part of the calculation when bringing this capability in-house. The more important test is whether it makes the data more useful to carriers.
