The future of insurance – the hopes and fears of AI

If we look in the rearview mirror, the last few years have really not been particularly good for the insurance sector. The fact that combined ratios are over 100% means that the underwriting performance has not been good and that businesses are effectively making a loss purely from the underwriting and the premiums that they are collecting.

Of course, there is also investment income to consider as well, which is something that companies have been relying on. However, the fact that interest rates have been quite low for a very long time is also something that undermines a long-term and sustainable business. These types of strategic effects is where the whole conversation has been going in the insurance industry for quite some time.

So, what is it that the companies can do to have an impact on the main drivers of their business; to firstly lower costs and reduce claims losses, whilst at the same time improving premiums, in a way that actually leads to sustainable business?

Ultimately, the insurance market is a very competitive industry, particularly in personal lines. It is predominantly now driven through price comparison websites and that means the pure price is up there front and centre for individuals to compare and choose from. This has been one of the biggest challenges that the industry has faced for quite some time. However, this approach can also lead to new opportunities, especially with the advance of technological innovations that are happening in the market.

Embracing technology

Rapid technology innovation certainly provides a number of levers that firms can utilise to create a competitive advantage, and this is where many are focusing right now. How can they use their vast data-sets to best advantage? How can they utilise artificial intelligence (AI)? How can they use technology to improve underwriting performance, have better pricing, greater micro-segmentation, and hyper-personalisation for customers, so they actually feel they are valued and not just a data point amongst a wide population?

Reducing claim losses

The second area of focus is to look at ways to reduce claim losses. Whilst end-user insurance fraud has been a factor in insurance since the industry was first formed, claims loss fraud is a growing concern for all firms. In particular, the rise of AI is actually leading to further risk of more fraud, since now it is becoming so sophisticated it can be used as a tool to make false damage claims.

Technology costs

For many firms, the rush to embrace new technology is tempting, however, increased costs can often be a huge challenge. Operational costs have been increasing on an annual basis and inflation has been a big problem for quite some time. Whilst the effect on consumers of the ‘cost of living crisis’ has been difficult, the overall inflationary driver has also had a big impact on the profitability of the insurance sector.

However, new technology does also allow companies to create efficiencies and improve productivity, by utilising more automation without necessarily reducing the workforce. It also allows for the efficient deployment of the most skilled workforce directly where most value is created, thereby avoiding costs in the longer term.

The way ahead

Looking forward, the priority for the next 12 months for most insurance firms is likely to be a continuation of the drive towards AI. Businesses across all sectors are talking about or experimenting with AI, yet very few are working with it in production or at scale – and yet everyone realises its importance for the future. AI is here to stay and for those working in the insurance sector they need to explore and begin to utilise the technology in the best way they can.

From an efficiency perspective, the best AI use-cases can really help to reduce cost through automation. However, more than this, it should also enable insurance companies to serve commercial and other specialist insurance markets where they might traditionally lose out, with inefficient processes that may take a week or even a month to close a deal. We have witnessed the experience of consumer price comparison websites that have made it easier for personal lines insurers to engage with customers; the opportunity now exists for commercial lines to make similar advances.

Of course, commercial lines and the speciality insurance space are not so paper-driven; there is a much greater focus on the detail of the specific risk and coverage, but this is definitely something that can still benefit enormously from the application of AI and large data analysis. This is where we see the immediate opportunity and for the medium-term.

Beyond AI, the other main focus for insurers is that of widescale digital transformation and the continued move towards cloud. Whilst some firms are advancing rapidly, there are still many lines of business within insurance (particularly in the commercial space, for example), where digital can be better utilised. This is both from the engagement perspective and gathering better data from customers, whilst also delivering improved services for end-user customers as well.

The final aspect of the future of insurance is of course regulation, which continues to drive change in what is a highly regulated market, as firms continually race to comply. However, going beyond mere regulatory compliance, the challenge is actually more about overall corporate resilience, especially in an increasingly complex technological world. The recent experience we witnessed with the unexpected widescale CrowdStrike disruption, was caused by a simple update in software actually intended to protect against crashes and disruptions. This example highlights the importance of firms having a thorough understanding of their entire and interconnected technology estate, in order to mitigate against increasing levels of critical vulnerability. For insurance firms, the unthinkable situation could be where a customer needs to file a claim immediately because of a sudden emergency; if the insurers claims system was ‘down’ or the data was lost in an outage, how would they manage?

Strategic data transformation

To be able to create a robust AI-led strategy, firms first need to solve the data fundamentals, define and create a thorough and future-proof data transformation strategy. Organisations need to map out a proper, available and reliable dataset which can also act as a single source of truth. Having this in place is key to enable the formulation of a comprehensive AI strategy.

Once the data strategy is defined, and a pathway for data transformation is enabled, the plan for utilising AI in all appropriate areas of the business can then be implemented, in order to realise the efficiency and growth benefits. If not, then any move towards AI may fail to deliver the anticipated benefits, as the programme remains isolated as an expensive and potentially risky proof-of-concept.