The insurance market is fully embracing artificial intelligence (AI) in a bid to drive cost efficiencies, according to Craig Beattie, senior analyst at Celent, the research and advisory firm.
There are three key factors leading insurers to utilise AI, said Beattie, at FinTech Connect: “An increase in processing power, an increase in data availability – not just in surfacing data within the insurer, but also in absorbing external data, and also algorithm improvements.
“These days these things are much more efficient and we’re able to rely on much better hardware as well,” he said.
Celent recently surveyed insurance market participants on the subject of AI and found respondents have identified potential impact of the technology across the supply chain.
Claims management stood out as the area most respondents (69%) see a high impact, ahead of customer service (56%), and enterprise risk (44%). Project development (39%), policy/plan administration (37%) and sales & distribution (34%) followed closely behind.
Beattie pointed out that AI has been around for eight decades, but developments in the technology come in waves. The most recent surge in interest from financial services has seen insurers look to embrace productivity, he said.
“Within insurance, we’re seeing two key drivers. We’re seeing cognitive and deep learning investments making it much easier for data scientists to be much more productive.," said Beattie.
“Secondly, there’s been a great democratisation of machine learning. Now it’s become so much easier for someone who doesn’t have a background in machine learning, doesn’t have a degree in mathematics or even in actuary, to start to leverage some of these AI applications. So these people can start with some data, throw it at something, and it generates AI application. It might not be as good as something that someone with three PhDs in physics might generate, but its going to be good enough to make a difference to the organisation and potentially across the insurance industry.”
Further, insurance firms are attempting to “industrialise” machine learning and AI – scale up applications and developments they have successfully implemented – according to Beattie. Further, the recent opening up of application programme interfaces (APIs), now allows firms to share knowledge, which is a further sign of the democratisation of the fintech sector he said.
“It’s so easy now for a developer to integrate through an API to something that’s going to use machine learning to improve or optimise an insurer’s competitiveness,” said Beattie.
In the same survey, respondents were asked which application of AI technologies were most prominent in each sector. Respondents said data science holds the most prominent AI application within product development (51%), followed closely by enterprise risk and marketing & branding (both 48%), and sales & distribution (43%). Robotic process automation (RPA) is gaining most traction for 40% of respondents in both claims management and policy/plan administration, ahead of back office functions (31%), and customer service (29%).