Over 70% harbour deep concerns around factors like brand damage, runaway risk from autonomous systems, and data errors propagating via generative models.
Nearly 90% of leading UK financial institutions have launched numerous AI solutions over the last year, reveals new data from EXL Services. However, many companies are investing heavily without first optimising their data architecture to support AI.
This oversight is creating risk exposures amidst rapid adoption of cutting-edge innovations like generative AI.
The findings come from EXL’s 2024 Enterprise AI Study covering the UK’s top 100 insurers, and leading 20 banks/alternative lenders.
Most organisations report double-digit million pound AI investments in the past fiscal year, with over 44% having introduced AI across 8+ business lines. Marketing, compliance, claims processing are major focus areas.
Naturally, this widespread implementation is facilitated by a strong willingness to invest in AI.
Almost 9 in 10 (86%) financial services (FS) leaders reported their firms investing upwards of £7.9 million ($10 million) in their most recently completed fiscal year, and 35% report investment of £39 million ($50 million) or more.
Despite bullish AI activity, under-developed data operations threaten successful roll-outs. Almost half of respondents (47%) admit their firms are only “minimally data driven” currently.
“The risk is that mounting competitive pressure can lead to AI investment that isn’t properly thought through,” cautioned Kshitij Jain, EMEA Practice Head & Chief Strategy Officer for Analytics at EXL.
“Getting data architecture and governance right upfront greatly enhances AI outcomes, and neglecting it can waste budgets and elevate risks.”
The research also highlighted a cohort of ‘Strivers’, representing 45% of respondents who report they are more narrowly implementing AI, across on average, four business functions.
However, the study suggests their implementation is deeper and more focused in many cases, and as a consequence there are areas where they are over indexing– particularly the use of AI to lower costs, where they are 23 percentage points ahead of the Leader cohort.
Most concerning is rapid generative AI adoption amidst lack of safeguards. Over 70% harbour deep concerns around factors like brand damage, runaway risk from autonomous systems, and data errors propagating via generative models.
Without addressing them, benefit expectations may prove unrealistic.
“Enterprise-wide AI success requires great strategy, governance and employee training – not just flashy technology,” Jain insisted.
“Firms properly identifying priorities, planning steady integration on robust data foundations, and mitigating risks will pull ahead as the pack thins out.”
Research is based on a qualitative survey of 64 C-level executives, vice presidents and directors engaged in strategy, technology and business process at the UK’s top 100 insurance carriers and top 20 bank and non-bank lenders, conducted online and by phone.