The COVID-19 pandemic and lockdowns accelerated digitization, tipping the scales in favour of insurers able and willing to place a decisive bet on technologies such as artificial intelligence (AI)—but this was not without risks, industry leaders told a SIRC panel titled “The Potential of Data Analytics and AI in Insurance”.

The surge in data interaction brought all the sought-after benefits of digital channels while rendering a new quality of data collected, including from parts of the population not previously visible in the digital world.

“The major question will be how much of this will remain after COVID-19,” said Fabian Winter, the chief data officer at Munich Re. “That will be a very interesting question.”

A more fundamental question could be how well the insurance industry is positioned to capitalise. “The insurance industry in total is not at the forefront of digitisation and AI, and the inroads made to date appear quite heterogeneous,” Winter said of the view from Europe.

“Maybe not every insurance company is completely convinced.”

Examples of AI in cross-selling, upselling and sale analytics may not be uncommon. But deeper use of AI across insurance operations remains rare, he said.

A full-scale set of investments can be daunting. Winter described a multi-faceted endeavour at his firm, requiring onboarding massive amounts of new skills and know-how and then integrating with internal expertise on insurance.

For every new technology or new skill entering the building, a host of existing staff need training for its use across the company and at varied levels of specialisation, he noted.

“It is important just to get started,” Winter advised industry colleagues. A small company size and an early starting point are manageable hurdles, more easily overcome by joining an ecosystem to help find and perhaps jointly craft the mechanisms.

More responsibility

With greater data comes greater responsibility, participants warned. The panel’s sole regulator, Prapapas Kulpawaropas, a senior official at the Office of Insurance Commission, the Thai regulatory authority, stepped forward to speak of rules and compliance. And at least one industry representative sees serious issues well beyond the laws in deeper ethical traps.

The rush into big data and AI put Thailand on a visible path to “making insurance easier, convenient and cheaper,” said Kulpawaropas.

“But every development will use more and more private data in the future, bringing a need to make sure that the companies handle the data properly.”

In Thailand, she noted, the drive into digitisation has truly picked up pace. Her regulatory agenda is a madcap run through the dozens of angles where digitisation drives growth, data analytics, product development, embedded insurance products and much more.

A lot can go wrong along the way when stirring up large doses of client data, Julie Batch, an executive in direct insurance at IAG in Australia noted.

“Data isn’t pure data,” she noted. Go back a few years in the life insurance book and underwriting decisions can include a hefty dose of judgement (read: bias).

“Teach your algorithms to mix old data with new and you can end up with some really terrible outcomes,” she added.

The time is now

The momentum gained in digitisation makes today a pivotal time for companies to dive into AI and digital channels or risk being left behind, Rosaline Koo, chief executive officer and founder of employee benefits insurtech CXA Group, told the panel.

She sees a world splitting into the tech true believers and the tech/AI agnostics, and the latter won’t make it into the industry’s pending incarnation.

“It has to be top-down, it has to be holistic,” Koo said of her firm’s work to transform insurers to date. “The C-suite must have a vision to digitise end to end.”

Half-steps towards a corporate digital revolution create more hassle than benefit, she argued. “With some insurers, it is just digital lipstick; it doesn’t connect to anything.”

Sales coming through a digital channel hit the same old reality: “just a bunch of humans scrambling and having bad service”.

The paradigm shift comes as consumers have learned to expect more, Koo indicated. Consumers with options at their fingertips are accustomed to personalised solutions to fit their life stage and needs, not the “here is my product, who shall I sell it to?” approach still dominant among insurers, she said.

At CXA group, a base offer in employee benefits leads to tailored upsell in life and savings and product insurance, she claimed.

Panellists proved mindful of limits to a pure tech channels and AI processes. In India, demographics and geography make varied demands in so large a market and insurers have to take steps to avoid alienating clients and partners, Parthanil Ghosh, head of automotive for HDFC Ergo General Insurance Company in India, told the panel.

“Yes, we are in the people business,” James Park, head of Singapore and Southeast Asia for Munich Re, concurred.

“These are people at the end we are serving.”

Main image: Shutterstock / GarryKillian