It seems like only yesterday when everyone was talking about the impact of Big Data on the insurance industry. Talk about Big Data now and you will almost seem old-fashioned – it’s all about artificial intelligence and InsurTech.

Artificial intelligence (AI) will soon be everywhere. It is making decisions about what we buy, when we buy it and how much it costs. It is controlling interactions between customers and suppliers. It is even talking to customers. The insurance industry is facing huge changes as AI steps boldly into every aspect of its internal operations and external relationships wearing the bright new clothes of InsurTech.

To some this seems like a revolution with all the instability and upheaval that it brings in its wake. In reality it is part of a rapid evolution of which Big Data was a crucial phase. It has brought new players into the insurance market with some, like Lemonade, the world’s first peer-to-peer insurance carrier powered by AI and behavioural economics, experiencing phenomenal growth over a very short time.

It is estimated that around £1.32 billion was invested globally in the InsurTech arena in 2016, up 32% on the previous year. The lion’s share of this was in the United States but the UK and Europe are increasing their investment (see chart below).

Incumbents or disruptors?

By no means all of this investment has gone into start-ups – the disruptors that many commentators claim are going to change everything we know about insurance. A significant proportion of that investment has been made by established players in innovations of their own.

“It is on everyone’s agenda, but how it plays out will vary enormously,” says Mathew Rutter, Insurance Advisory Partner at DAC Beachcroft, “and no one quite knows where the smart money should be going and what is going to take off. Large insurers have created digital garages to accelerate development and that seems to be working well. But there are also plenty of small insurance start-ups that could develop niche products. Their lack of size might enable them to move very quickly.”

He believes the huge importance of AI and associated technologies was vividly illustrated at the British Insurance Brokers’ Association conference in Manchester earlier this year: “One of the striking things at the BIBA conference was the sheer number of stands taken up by a wide variety of technology exhibitors”.

There are differing views on both sides of the Atlantic, but many experts feel established players have as much chance of success as start-ups.

“The biggest winners from InsurTech may be those innovative incumbents who find better ways to underwrite through better use of technology. P&C [property and casualty] insurance is characterised by a wide spread of returns between winners and losers, and over time the best underwriters displace the rest of an otherwise lousy market,” said Adrian Jones, Head of Strategy and Development at SCOR Global P&C in a recent article.

“Trying to stand aside from this rapid evolution is simply not an option,” says Rutter. “Part of the issue for insurers is that even if they don’t invest and adapt they will need to respond to the technologically-driven changes around them, such as driverless cars. It is hard to see old business models being fit for this new world.”

Rhiannon Webster, Head of Information Law at DAC Beachcroft, agrees that it will be impossible to ignore the new world of increased connectivity and constant data flows: “The internet of things is going to generate a huge amount of new data. It will require a step change in the capability of insurers’ systems.”

Widespread innovation

The InsurTech label tends to be slapped onto anything that looks innovative and has technology and data at its heart, but it takes many different forms and some of them won’t look that exciting to the outside world. “It might just be a more efficient claims process,” continues Webster. “The customer may not see the technology, but will be better served. Smarter use of technology and data will enable a smoother process to be put in place for the customer, while for the insurer it will mean lower costs.”

Other innovations, such as fractional insurance where customers buy on a pay-as-you-go basis or peer-to-peer insurance, will have a deeper impact. “These innovations will change the risk and a lot of focus is on the consequences of that.”

Embracing those changes and their consequences requires a major shift in the culture of the industry, according to Webster. “This isn’t an entirely new challenge but insurers are approaching it in different ways. Some have had trips out to Silicon Valley to see what they might invest in because they want to find new thinking and new ways of working.

Many of them will invest and develop these as a separate business. “This isn’t a new approach. HSBC did it with First Direct which is operationally a separate business. Others might want to integrate the new way of thinking into their business, such as Aviva’s ‘no questions’ approach to home insurance.”

Webster argues that no one can say one approach is preferable to the other: “It might be better to let new ventures grow separately and develop the new systems required to cope with the vast amounts of new data and new ways of working”.

Overcoming challenges

For Rutter, one of the key cultural challenges for the insurance industry is going to be its cautious approach to regulation. “Regulation is usually designed to be tech-neutral, but insurers will be looking for reassurance. It is a lot to do with culture. Many organisations have a low-risk or no-risk approach to regulatory compliance. Anything that pushes the boundaries might ring alarm bells.”

The Financial Conduct Authority is the lead regulator in this area and it has been trying to engage the industry, setting up a ‘sandbox’ to encourage insurers to work with it to explore the impact of regulation on technological innovation. In particular, it will be aiming to test the boundaries of legislation such as the Insurance Distribution Directive (IDD). This could present problems.

“The IDD still largely assumes that most insurance is sold through brokers in face-to-face contact,” says Rutter. “It does have a bit of a feel of being where the UK was ten years ago. One potential effect of Brexit is that the FCA could be more flexible in varying or waiving rules that might inhibit innovation.”

He sees regulation as having an important role to play and says it need not be a negative influence or an inhibitor of innovation. “The FCA also has responsibilities to promote competition and the new entrants could improve consumer choice. There may be regulatory risk but there may also be regulatory benefit.”

Those bringing new insurance offerings to the market see the challenge of regulation in a different light, according to Dylan Bourguignon, Chief Executive of so-sure, a social insurer launched last year, writing in a recent article first published in FinTech Times: “Insurance regulation is even tougher than banks. In the lending market, there was a regulatory loophole with simple loans which were not regulated as parent-child loans existed before the banks did. This loophole led to the emergence of peer-to-peer lending platforms, as an alternative to bank loans.

“In insurance, there was no such thing as ‘simple insurance’ before Edward Lloyd’s coffee shop in the 1680s, so every insurance product sold in the UK is regulated. While it is required (as it protects the consumers), this regulation is such that if one tries to save money bypassing segments of the traditional value chain, the costs are such that they ‘kill’ the economics. So ntrepreneurs have to work with the industry unless they recreate the whole value chain.”

Data risks

One area of regulatory risk that will definitely have to be kept to the fore is data and privacy, especially when working with partners in fragmented supply chains, many of whom may not be familiar with data protection regulation, especially the new European General Data Protection Regulation (GDPR). This will be implemented in the UK by the proposed Data Protection Bill announced in the Queen’s Speech, with additional privacy law requirements.

“Many of the start-ups and innovators are American and they have designed their processes to conform to US privacy law which is not the same,” says Webster. “In Europe you will now have to build in to your systems privacy protections that are compliant with the GDPR. The US approach may not be easy to translate to the UK and Europe. In particular, internet scraping for data is very common in the US and there is just no easy way that can be made to conform with European data protection rules. This isn’t a question of who owns the data. In Europe the key issue is the protection of the individual’s personal data, whatever the source.”

The new rights given to individuals to check and control how their personal data is used are extensive, including the right to be opted out of automated processes. All of this has to be factored in at the product and process design stage.

Regulatory responsibilities will reach right up to board level. “Boards will have to ask probing questions about what checks and balances are built into the new AI-driven systems and ensure that the quality of management information is appropriate. Saying you don’t understand how it works won’t wash with regulators. This happened with the wholesale banks and algorithmic trading – many at board level were not sufficiently able to assess the risks.”

There will be some InsurTech applications that get it wrong, predicts Rutter, potentially selling large numbers of policies to the very people underwriters don’t want on their books: “Insurers need to understand that once automated decisions have been made, you can’t pull back from them by cancelling policies. That is hardly treating customers fairly”.

There will be failures and mistakes amid the whirlwind of innovation, but that isn’t a good reason for standing aside: “Often the only way to find out what is going to be best for a business and its customers is to try it,” says Webster.