As AI makes inroads into reinsurance, some fear it could erode jobs. But Ben Rose (pictured), co-founder and president of reinsurance platform Supercede, believes the opposite is true.
“There are probably less than a thousand people in the whole global reinsurance market who actually understand the nuances of buying reinsurance,” he said. “And that’s probably gotten worse, if anything.”
Rose said the growing burden of administration and the lack of investment in buyer training has left a shrinking cohort of experts handling some of the industry’s most complex decisions.
“I think probably one of the best things to look at here is something called the Dunning-Kruger effect,” Rose said. “The more you learn, the less you know.”
“In reinsurance, we’ve been quite expensively hedging our way along through very cautious overspend and uncertainty loading. But because we’ve got a very deeply flawed data exchange process, the industry’s kind of been operating from a position of accepted ignorance for a long, long time.”
That’s where Rose sees real opportunity for AI, not as a replacement, but as a catalyst.
“With stronger data foundations, AI can really offer us the chance to blow the doors off that old model,” he said. “Uncovering that ignorance and seeing how much there is out there to learn is going to create a lot more work and a lot more interesting work for everybody in the industry and beyond.”
While graduate pipelines exist for brokers, underwriters, and claims professionals, reinsurance buyers follow a different path - if they’re lucky enough to find one at all.
“Buying is the least professionalised career journey,” Rose said. “You fall into it and learn by following the boss or the boss’s boss.”
This model, he warned, is reaching a breaking point. “Old school reinsurance buyers are already starting to suffer a bit. They’re turning up still with their colour-coded spreadsheets and lots of nostalgia about long-standing relationships, but that’s not working so well.”
“As they retire, there’s going to be a very serious drop-off and collapse in understanding of how reinsurance buying should work in these newer generations.”
For brokers, that has implications. Inexperienced buying teams can stall placements, default to “safe” structures, or resist innovation due to lack of confidence or capacity.
Rather than displacing talent, Rose believes digital platforms can restore it.
“We’ll just buy the same thing that we bought last year because we don’t have time to consider anything else,” he said. “Or it’s too complicated to change things just in case we change to something that might be worse.”
He sees that changing with better systems in place. “You could end up with much more nimble reinsurance purchasing teams,” he said. “Aided by strong data foundations and possibly AI in due course, reinsurance buying can actually become much more precise and effective.”
“Maybe we’ll see even more reinsurance being purchased more frequently. We’ll get out of these big annual cycles as well. Who knows?”
While large reinsurers and brokers continue to attract graduate interest, smaller buyers often struggle to compete for talent, especially when the job involves hours spent managing inconsistent spreadsheets.
“Unless you’re the top dog who only handles relationships and doesn’t actually have to do any of the data work, you’re essentially signing up to death by spreadsheets,” Rose said.
That dynamic, he added, is fuelling high turnover. “There’s a merry-go-round of talent because there’s not enough people coming in. The people who are in are unhappy and have quite a high dropout rate.”
He said some firms are now gaining an edge by modernising the buyer experience.
“Being able to offer their team members or their potential recruits the latest and greatest reinsurance buying software to get them out of the weeds has actually given them an edge in that war for talent,” Rose said.
With major technology firms leading the AI race, Rose believes insurers and brokers should avoid trying to build their own competing systems, and instead focus on readiness.
“You don’t need to go and build it in-house - you’re going to get access to these tools very, very soon,” he said.
Rather than duplicating efforts, firms should be investing in data quality, process consistency, and digital workflows, steps that will enable AI adoption when it becomes practical.
“Actually we should be getting our house in order,” Rose said. “Because when that AI moment comes, we’re going to need to be ready to play ball.”
