Greater focus on data optimisation needed to curb ‘costly’ AI implementation mistakes

Data optimisation must be a ”greater priority” for UK insurance firms when implementing artificial intelligence (AI) so they can avoid making “costly mistakes”.

That was according to data analytics and digital operations and solutions company EXL, which said in a statement yesterday (29 January 2024) that more financial services firms have invested millions of pounds to implement AI.

The firm’s research found that 86% of firms had invested upwards of £7.9m in the most recently completed fiscal year, with a further 35% investing £39m or more.

It also revealed that 89% of firms in the UK had launched AI point solutions and proofs of concept over the past year.

However, with the “mounting pressure” that firms are facing to implement the technology, “the risk” is that this “can lead to investment that isn’t properly thought through”, Kshitij Jain, EMEA practice head and global chief strategy officer at EXL, said.

“A need to move quickly can mean ensuring operations are truly data driven gets deprioritised and this can be a costly mistake,” he added.

Data architecture

The research was based on a survey of c-suites and technology experts at the UK’s top 100 insurance carriers and top 20 bank and non-bank lenders.  

The firm also found that 44% of firms had already implemented AI across eight or more business functions – the most extensive areas where it was deployed was in marketing, business development and regulatory compliance.

But despite the level of investment, some 47% of respondents believed their firm was “minimally data driven”.

Jain said “the outputs and impact of AI” can be “greatly enhanced when implemented within an organisation that puts data architecture, quality and governance at the top of the list”.

“From AI-driven back-office processes to customer-facing generative AI content, the key with any successful AI roll out is a measured and strategic approach,” he added.

“Getting data architecture right, experimenting with solutions in a sandbox environment and training employees properly are critical steps in the journey – neglecting them can result in wasted investment and greater exposure to risk.”

Generative AI

Looking at generative AI, nearly three quarters of UK firms were found to already be using the technology, compared to 50% of US counterparts.

And over half also admitted to investing more in AI, due to the growth of generative AI.

However, some 70% of UK firms reported being concerned about the use of generative AI – citing fears such as its impact on a firm’s brand in the case of its use going wrong (41%), risk to organisational brand reputation (36%), generative AI operating anonymously (34%) and inaccurate impacting outcomes (33%).

Jain said that for enterprise-wide adoption to succeed, ”boards must be bought into AI’s capabilities and AI must be tightly linked with strategy”.

“Whilst it’s brilliant to see the evident enthusiasm for the technology coming from leaders at the UK’s top insurance and banking firms, the factor separating the wheat from the chaff now is how effectively investment is actually being used and ensuring that implementation is wide, yet focused,” he added.

“Those firms able to properly identify priorities and plan phased and growing AI implementation, whilst ensuring they are appropriately data driven as a critical foundation, will find themselves pulling away from the pack.”