Tim Grafton, chief executive of the Insurance Council of New Zealand, joins Insurance Business with a new monthly column.
The Financial Markets Authority will shortly grant its first exemption to permit Artificial Intelligence (AI) to provide financial advice. Current law requires financial advisors to be natural persons, hence the exemption before new law formalises matters.
AI will revolutionise customer experiences beyond the provision of robo advice. Data is central to that, supported by the speed of automated analysis built on algorithms programmed by humans. Customers will increasingly benefit from the speed of claims settlements. They will enjoy the convenience of obtaining the right amount of cover tailored to their individual needs. Better, value-added services will be available that enhance and evolve the customer experience beyond being a pure insurance product.
Insurers will be better equipped to provide advice to customers to reduce their exposure to risks and hence reduce costs for both. They will have granular detail that will inform their appetite for risk and how to price it. Underwriting may become more scenario-based, predictive and less reliant on historical data, especially if we are talking about new risks. More functions will be automated releasing staff to focus on adding value to the customer experience.
Mobile platforms will become the ubiquitous delivery channel. Online ecosystems will integrate multiple customer needs into one experience. Insurers will be attracted to partner with other sectors to grow, deliver better value and meet what will be standard consumer expectations.
Technological change of itself will reduce some risks and create new ones – think of the safety of autonomous vehicles and the exponential increase in cyber risks arising from the growth of the Internet of Things.
It will be a more complex but rewarding environment to operate in for those who adapt and partner well. Inevitably with major change there will be casualties too. The profile of the insurance sector will undergo fundamental transformation with the customer at the core of that change. Insurers will need to be fast, low cost and offer relevant services.
Inherent to insurance from its inception as a form of mutual protection to whatever the future holds is trust. You can’t base a business on seeking payment in return for a promise to restore the possibility of future loss without it.
Trust is hard won and easily lost. So, how could trust be damaged?
Decisions made by AI or Big Data analytics, can be flawed or even biased. AI is based on identifying clusters of patterns in data, but not all clusters are meaningful or relevant. No algorithm yet reconciles moral values and self-interest, so human judgement is still required.
Consumers will not want to feel their privacy has been violated by the data collected about them and how it is used. Insurers will need to be clear about what data they do rely on to gauge risk, transparent about the data they are accessing and the purpose to which the data is being put, and perhaps, just as importantly what it is not being used for. This will invite interesting conversations between insurers, regulators and consumer interests.
The situation is made opaque by the accessibility and the ambiguity of data ownership, such as, whether a public post to the world on Facebook should be used to assess an individual’s risk.
It will not always be easy to know at what point consumer consent is required. Consider, how different databases are sourced and aggregated and analysed to profile an individual’s risk or where that aggregation occurs in an ecosystem involving multiple partners providing an individual consumer with a customised product.
Data-mining firms will set up shop, likely domiciled in jurisdictions beyond the shadow of New Zealand regulators, alongside others providing products into our insurance market.
If technology enables insurers to accurately cherry-pick the risk of individuals, as opposed to whole populations, what impact will that have on consumer trust and traditional business models?
Harnessing Big Data to benefit customers will require high levels of transparency about the use of data, as well as ensuring the integrity of the new tools, to preserve public trust and the social license that confers. This will enable delivery of the benefits consumers demand - value for money, convenience, speed, efficiency, competition and consumer-friendly solutions.
Related stories:
“The one thing a machine doesn’t have is empathy, and that’s what people are looking for”
Proud year for the insurance sector, says ICNZ chief
