The AI revolution: The pace of change accelerates
The AI revolution is underway. We are already living through an era of profound change that will be talked about for generations to come. Just as the 1780s are talked about as the start of the Industrial Revolution, so the 2020s will be seen as the dawn of the ‘AI Revolution’.
Reading certain headlines, you would be forgiven for thinking that use of AI by the insurance sector is a new phenomenon. The reality is that the sector has been integrating elements of AI into the insurance lifecycle for many years, be that customer service chatbots, fraud detection tools, telematics, personalised marketing, risk assessment and underwriting, predictive analytics for trend forecasting or automation of back office functions. These examples have ultimately brought about more sophisticated product offerings, reduced claims management times and increased risk management – all to the benefit of the policyholder. However, the pace of change has undeniably accelerated over the last 18 months.
The explosion of generative AI – both in terms of technological advances and general public awareness - during the summer of 2023 was a pivotal moment.
“Since then, many organisations have found themselves in an ‘AI sprint’, racing to determine appropriate use cases and strategy. They are keen to demonstrate engagement and advances to clients and investors and to stay ahead of the competition, but, at the same time, they are also having to work through and fully understand and mitigate the risks,” says Jade Kowalski, a DAC Beachcroft partner specialising in data and privacy. AI has quickly become part of the day-to-day conversation at all levels, from the boardroom to the water cooler. Picking a path through the AI labyrinth requires a clear sense of purpose, says Simon Konsta, a commercial disputes and insurance partner at DAC Beachcroft: “Organisations need to be very clear on their strategic priorities and value expectations and share that vision with their client base, their staff and teams. Clear sighted governance, robust risk management and technical oversight are absolutely essential with a technology that is evolving quickly. “This means oversight of the development of the technology, oversight of the use cases and implementation of appropriate safeguards.”
The regulatory environment
If keeping abreast of the rapidly expanding capabilities of AI is not enough, the complex regulatory environment surrounding the use of AI, which includes existing laws and regulations such as the GDPR and other privacy laws, intellectual property and copyright issues, consumer protection and product liability, is also crucial, says Kowalski:
“In the EU, the AI Act has now been passed. However, that is very much the beginning of the journey. We are expecting a plethora of guidance, both at EU and Member State level, and it will be some time before the concepts that we are now grappling with transform into a fully developed regulatory framework; and longer still until we have market practice and conventions.
“In the UK, we are at an even earlier stage in that journey. The current approach is for existing regulators (such as the Information Commissioner’s Office) to regulate by reference to principles. Although, of course, that may change, particularly if there is a change of government.”
AI risk frameworks
The drive for the insurance sector to use AI, set against a complex regulatory background, brings into play a wide range of risks, resulting in the need to put in place and embed appropriate AI governance processes and systems. However, it takes time to develop such processes into a comprehensive risk management framework; and longer still to properly embed them.
“At the same time as in-house legal, privacy and compliance teams are looking to set up these comprehensive frameworks, they are also doing their best to support and facilitate the plans of business colleagues who are raring to go with new AI initiatives. Juggling the two competing demands and pacing activities at both ends of the spectrum can be a challenge,” says Kowalski. The buy in of key stakeholders and the upskilling of colleagues across the board will be vital if an organisation wants to integrate AI successfully into its business.
Impact of AI on underwriting and claims exposure
Internal use of AI by the insurance sector is just one element of a much bigger picture. The impact of AI on the provision of insurance is also under scrutiny.
“The landscape here is also developing,” says Konsta. For certain classes of business, such as tech E&O, recent advances in AI haven’t come as a big surprise; they are part of the evolving nature of the technological world. However, in other classes, the pace of change has caught some insurers (and their policyholders) by surprise.
“I think it’s fair to say that the market is still grappling with the impact of AI on professional indemnity risks,” says Catrin Davies, a partner and professional indemnity specialist at DAC Beachcroft. Professions such as accountants and lawyers are increasingly using AI for all sorts of different tasks, including preparing draft reports; summarising lengthy meeting notes; research; and crafting marketing content. “None of this is without its risks and having a clear set of protocols, backed by staff training, will be essential if firms are to avoid being drawn into disputes that could test the boundaries of professional indemnity policies, which are still evolving when it comes to the risks presented by AI,” says Davies.
“There are also all sorts of risks surrounding bias and hallucinations that require constant vigilance. This is not theoretical. Those piloting the use of AI are seeing real life examples of hallucination,” says Konsta. “It is a critical certainty that AI will become part of a new normal for professional services firms but it should be clear where AI stops and skilled professionals step in.”
So how will the insurance sector adapt to meet this challenge? “Ultimately, the insurance sector is built on understanding evolving risk; both its own corporate risks and those faced by its policyholders and I have no doubt that insurers, brokers and others will continue to raise the bar when it comes to managing AI risk,” says Konsta.
