Are insurance regulators turning their eye to data ethics?

Insurers in Australia and New Zealand, like insurers across the world, continue to invest heavily in data and analytics to drive business outcomes. Risk selection - pricing and underwriting - has attracted a lot of attention. But this energy also goes into everything from marketing to claims - eg fraud.

This drive to find more ways "to make money" and "serve customers" has sometimes lead to some less than desirable practices, and sometimes some "accidents". The UK Financial Conduct Authority (FCA) has recently introduced rules to tackle “price walking” practices disadvantaging loyal customers. In Australia, several insurers have had issues with pricing/loyalty/discount structures.

There is no doubt that as more data becomes available, more compute power is leveraged and more techniques can be deployed (machine learning etc), insurers will continue to push to find competitive advantage through data - just as they have done in the past. With all this data, and these techniques, what is actually going on with data driven decisions can become very opaque. And the risk of bias is real....

Very interesting write up by Duncan Minty on recent statements by the regulator in California on this issue. Well worth a read.

Insurers would be well served thinking long and hard about how they ensure they have the right checks and balances in place. Left unchecked, the drive for results can see some other than desirable outcomes. Additionally, "unintended" bias can, and does happen!

My colleagues at the Gradient Institute are doing some great work in this area. I am sure Tiberio Caetano and Bill Simpson-Young would happy to share their insights and approaches if you are interested.

Source: Duncan Minty