Machine learning and artificial intelligence will play an important role in claims assessment but can never dislodge the need for human expertise, a leading data scientist said at a claims conference organised by Asia Insurance Review yesterday.
“I see machines being able to assess a large percentage of claims but they would be handling the standard types that allow for straight-through processing, and thus allow humans to focus on the remaining 20% of claims which are complex and will definitely require human expertise,” said Prof Shonali Krishnaswamy, Chief Technology Officer, AIDA Technologies.
While acknowledging that AI will likely result in lower headcount in claims, she added that AI actually refers to ‘Augmented Intelligence’ as it is meant to enhance the ability of humans to perform their jobs better rather than a switch to automation.
“The machine learns from interaction with humans and build predictive data and then identify outliers, but it is the human experts who are the final authority on how the claims should be settled,” she said.
“The use of machines will be a phased approach and human experts will have to continually input adjustments into the machine which it will recognise and learn, and over the long run there will be greater convergence but it can never totally replace human expertise,” she added.
The General Insurance Association of Singapore estimates that one in five claims the industry receives are either false or inflated, which works out to approximately S$140 million (US$101 million) each year.
Held over two days in Singapore, the Asia Conference on Claims Management and Insurance Fraud is sponsored by Singapore Re and ends today.