Bringing life insurance into the age of Big Data
Life insurance is one of the oldest industries in existence, but today it is undergoing a transformation thanks to AI and data science.
The big picture: With a business model built on predicting the future of its customers, the life insurance industry seems well-positioned to take advantage of the prognosticative powers of machine learning, but it will have to overcome an ingrained conservatism — and fears of AI bias.
By the numbers: A report published this week from the analytics company GlobalData forecasts that AI platform revenues in the insurance sector will grow by 23% a year between 2019 and 2024, reaching $3.4 billion.
- Startups like Lemonade — which last year became the first "insurtech" company to go public — have worked to overhaul the time-consuming process of buying insurance, with a digital-first interface and machine-learning analytics.
What they're saying: "What was not being used as data made no sense to us," says Paul Ford, who spent years in the conventional insurance industry before cofounding the insurtech company Traffk.
- While legacy life insurance providers use outdated actuarial tables to determine whether and what kind of policies to provide customers, insurtech companies like Traffk draw on thousands of data points to provide a more personalized analysis.
- "It's all about creating a picture of risk that is high definition, as opposed to what television looked like in the 1950s," says Ford.
What's next: A report put out last month by McKinsey envisions a near future where AI has shifted the industry from its current state of “detect and repair” to “predict and prevent," leading to active insurance products that respond to changing customer behaviors in real time.
Yes, but: Some experts worry more precise insurance rates set by AI could end up discriminating against certain groups, which in turn could draw the attention of regulators.