Uncertainty caused by unclear data adding up to 10% on cedents’ premiums

Cedents that submit unclear or inconsistent data to reinsurers when trying to obtain coverage could face a premium uptick of up to 10% thanks to “uncertainty loading”, according to Ben Rose, president at reinsurance technology platform firm Supercede.

Speaking to GR during RVS 2025 in Monte Carlo, Rose warned about an “inbuilt danger” that is impacting some primary market firms that are utilising artificial intelligence (AI).

He explained that if reinsurance underwriters “receive submission information” that “doesn’t make sense or looks inconsistent with last year”, then “the only safe thing [they] can do to respond is to charge a higher price – just in case”.

Rose described this situation as “uncertainty loading”, which he has seen add between 5% and 10% to cedents’ reinsurance premiums. He estimated that uncertainty loading can add around £25bn a year across sector-wide premiums, “just for not getting your data into order”.

He continued: “You have these clients [that] suffer on a reinsurance spend of a few £100m [because this figure] can grow by £10m – if not an extra £100m – just because their data was unclear or inconsistent.”

Rose told GR that “the biggest danger for the reinsurance space is that it may try to run before it can walk” with regards to its usage and implementation of AI because it is still “missing the key ingredient for good AI, which is solid data”.

He continued: “If [the market tries to] charge into the world of AI without a solid, reliable source of primary data – which at the moment, [it] often [does not] have – it is akin to building a house on very shaky foundations.

“That’s, unfortunately, the way the reinsurance industry is built – it depends on the data that it receives from the primary insurance market, which is often not very reliable or accurate due to the way that information is shared with the reinsurance industry.”

Siloed platforms

For Rose, this data inaccuracy stems from the fact that “historically, the insurance industry’s technology infrastructure has been built in 1000s of tiny islands”, with each company creating its own in-house technology that then struggles to communicate with another firm’s version of the same platform.

“They realise that none of their versions will talk to each other, none of the data structures or standards used match up or are the same and so they find themselves unable to communicate in an industry that depends on being able to communicate with partners,” he explained.

He acknowledged that this where his firm, Supercede, hopes to help – the reinsurance technology platform aims to get “clean data flowing into reinsurance companies” by acting as “a consistent bridge that gets information into the same format”, therefore avoiding “really expensive pricing loads for clients”.

Bullish about AI

Despite his plea for caution around AI being stacked on top of inadequate data, Rose admitted that he is “actually very bullish” about the opportunities this technology can bring to the market.

He commented: “It’s easy for us to have lots of opinions on AI being a scary thing, but I am actually very bullish about it.

“In the longer term, we can run. I know I said we shouldn’t run before we could walk, but if we can figure out walking, then there is a really bright future ahead for the AI world – but we need to focus on getting [walking] right first. And then we can unlock [an AI influenced future].”

To download the full Monte Carlo RVS 2025 annual issue of GR, click here.