There has been a lot of talk in recent years about the potential for AI to transform the insurance industry.

In this post we’ll discuss how AI is currently revolutionising the world of insurance, while also looking to the potential new ways insurers might use AI in the future.

What Type of AI Are Insurers Using?

Currently, when people talk about AI in insurance, they are largely talking about “narrow-AI”. That is, AI that only has a limited function, or which can only carry out a single task. Technologies include Machine Learning, Natural Language Processing, and Computer Vision tools.

Narrow-AI has the potential to radically transform the way insurers work:

• Automation – AI modules can automate certain repetitive tasks, such as filing.
• Data Processing – AI can quickly sort through immense datasets and automatically generate insights, which insurers can draw from when making decisions about risk assessment etc.
• Product Development – AI can automatically design, configure, test, and price insurance products to ensure they always meet the customer’s needs at the fairest price. For example, the parametric Flight Delay Compensation tool uses an AI model that can predict flight delays. In the event of a delay, it can provide affected customers with an instant pay-out with no need to make a claim.

The Current Limitations of AI in Insurance

As narrow-AI tools can only perform a single function, or a limited set of functions, they can only work as part of a wider system. In other words, they still need some human input. Insurers might turn to AI to process data, but they’ll still make the decisions about products, risk, and claims themselves.

So, currently, AI can add value to insurance only when it works in close combination with human processes.

There may be a future for generative-AI in insurance. While a narrow-AI tool is essentially a mathematical model capable of learning from data, a generative-AI tool is theoretically capable of thinking for itself. Examples include the ChatGPT “virtual assistant” and Google’s Gemini – systems which can respond to fresh prompts with almost “human-like” intelligence.

Generative-AI in Insurance – The Future?

Generative-AI is an emerging technology. No system has yet been able to quite replicate human thought or insight, and perhaps this will never happen.

AI User Case: Customer Experience

Yet we have already seen examples of insurers drawing from generative-AI models to improve the customer experience. KPMG, for example, recently announced that they have integrated generative-AI into their global smart audit platform.

Essentially, this platform provides auditors with intelligent virtual assistants, which can:

• Quickly review documents to instantly identify risk factors.
• Develop testing procedures to help auditors rapidly respond to emerging risks.
• Automatically summarise and document audits, including automatically reviewing financial statements.

AI User Case: Fraud Detection

KPMG are not the only major firm to experiment with generative-AI. EY recently revealed that they are using advanced AI systems in their auditing processes to automatically detect fraud.

Managing the Risks of AI

Just like any other technology, AI brings new risks. For example, will generative-AI models be subjected to the same stringent regulations that insurers must adhere to? Is it ethical to allow AI to make decisions about risk, cover, and claims?

Beyond this, there are also numerous challenges facing any insurer who wishes to experiment with AI. Currently, the biggest hurdles seem to be the cost and time of implementation, coupled with the lack of technical capability and skills.

Yet there is no doubt that AI has the potential to change the insurance world for the better. AI tools can automate numerous complex and time-consuming tasks, giving insurers more time to focus on delivering value to their customers. It can quickly analyse huge amounts of data, giving insurers quick access to the insights they need to respond to changing markets and emerging threats.

AI can also improve the customer experience, with innovative products, automated claims, and virtual assistant capable of responding promptly and accurately to any enquiry.

At Capacity Insights, we draw from predictive modelling, dynamic analytics, and other data science systems to drive our underwriting strategy, ensuring we can always deliver excellent value in an ever-evolving market. Find out more about how we use data science insights to create our bespoke insurance schemes.