Briefing: Generative AI in insurance is here to stay – and there’s no going back

By Clare Ruel

Generative artificial intelligence (AI) is here to stay in insurance – and even the UK government is on board with it.

This type of AI can produce new content, chat responses, synthetic data and deep fakes. It has been widely used in insurance in the form of chatbots, underwriting algorithms and claims due to its ability to analyse mass data sets with ease and at a minimal cost.

For example, Zurich has recently been experimenting with ChatGPT to overhaul its claims processing, while Lemonade was one of the first insurtechs to use it to pay claims in three seconds.

But now the government has fired a warning shot over the safe and ethical use of AI, alongside pumping an extra £100m more in investment into generative AI in a bid to make the UK “a technology and science superpower”.

In a government whitepaper entitled A pro-innovation approach to AI regulation, published on 29 March 2023, the government issued five new principles to regulators – including the FCA and PRA – to ensure the ethical practice of AI for insurance firms.

These included – safety, security and robustness; transparency and explainability; fairness; accountability and governance; and contestability and redress.

This desire for coordination between insurance regulators has been put in place to develop context-specific approaches to AI, rather than requiring the estalishment of a single AI regulator.

With this, a consultation has been launched which will run for 12 weeks until the 21 June 2023, with an AI regulation roadmap to be published alongside it.

But is regulating generative AI that simple? And will this regulation stifle innovation, especially in sectors like insurtech which need to be malleable?

An evolving regulatory challenge

I recently spoke to Andy Thornley, head of financial services at the trade body TechUK, who told me that “the FCA has previously consulted on how AI might be regulated”.

He continued: “TechUK’s position on this has been one aligned to outcomes, rather than a rules-based approach which makes regulating an evolving tech challenging.”

Meanwhile, Thornley pointed to a joint FCA and Bank of England survey of 168 firms published in October 2022, which looked at machine learning – a branch of AI. It revealed that the insurance sector saw the largest uptake of this technology across the UK.

Thornley added: “Therefore, the regulatory landscape ahead is perhaps more important to insurance.

”Shaping how this will be applied so that firms and customers alike can realise the benefits of this transformative technology should be on the agenda of brokers and insurers alike to ensure that the specific context that applies to them is accounted for in any future regulations.”

TechUK has had significant engagement with both government and regulators to help shape the government’s approach to AI and continues to work with policymakers as this develops.

With this in mind, it’s clear that AI holds a plethora of opportunities for insurtechs and for insurance overall, despite naysayers harping on the dangers of ChatGPT.

Italy became the first country to ban ChatGPT back in March, demanding its developer Open AI demonstrates how data was being used. More countries could follow, as the Italian government’s concern focuses on general data privacy regulations, which have been adopted across the European Union.

But, I’d like to think more in the terms of Wall-e than The Terminator when it comes to ChatGPT. Don’t insurance chatbots want to be your friend?

Personally, I believe AI offers masses of potential to the insurance market and the industry has only just begun to deploy it on a wider basis. It even has the ability to augment the human underwriter too – something which digital-first syndicate Ki has already cottoned on to.

Now with this new shot of government funding aimed at insurtechs and startups, the industry can push generative AI to its limit while keeping innovation alive – plus the extra funding won’t go amiss for the insurtech sector, as it has historically experienced problems gaining financial backing to incubate ideas.