Last November, Gen.Life introduced to DigFin its model of using machine learning and primary data to let an insurer generate benchmarks to price risk. The A.I. company is now preparing to partner with five insurers to form a joint-venture insurance company that will sell insurance online.
“The projects are in early negotiation stage,” said Steve Monaghan, chairman and CEO of Gen.Life. He declined to name the insurers but says they are well-known in Asia Pacific region. The joint venture will sell insurance under their licenses, with Gen.Life serving as a technology platform.
But the startup no longer has AIA as a partner. Both parties confirmed to DigFin that the Hong Kong-based insurer decided to end its research-and-development contract with Gen.Life, leaving its only partner American Family Insurance in the U.S. Monaghan says Gen.Life is still delivering a claims automation platform to AIA. “So there is still a technology relationship,” he said.
The initial focus of Gen.Life’s proposed J.V. will be life and health insurance, but will gradually extend into other types of insurance, Monaghan says. It is to be a B2C model, selling direct to consumers.
“Gen.Life will become the Spotify of the insurance industry,” said Monaghan, arguing that personalized and affordable insurance products can have the same impact in financial services as Spotify has had in music.
Gen.Life is a Hong Kong-based company using A.I. tools to learn from many sets of primary data, letting insurance partners use behavior data as benchmarks for pricing risk, rather than traditional actuarial calculations based on filed reports. The more carriers involved, the greater the pool of data, and so the more effective the A.I.
But as the AIA pullout shows, getting insurance companies to agree to share their data is a difficult sell – although it’s not clear whether AIA was wary of the business idea, or with Gen.Life in particular. Monaghan previously worked at AIA as its “head of edge”. He has also served as DBS’s chief innovation, in Citibank’s e-commerce business, and head of consumer finance at OCBC.
Monaghan argues primary data sets can help better underwriting and automate claims, while machine learning can improve the efficiency and quality over time. “We don’t change the risk profile,” he said, “but the more we understand risk the more we can manage it.”