Concerns raised over ChatGPT use in underwriting experiments

Firms have been issued a warning that they could experience “difficulty” with ChatGPT when using it in experiments to help aid underwriting.

Piotr Piekos, insurance lead at technology consultancy and software delivery partner Future Processing, highlighted that the tool can be “unreliable” during a roundtable chaired by Insurance Times technology editor Clare Ruel yesterday (7 June 2023).

Panellists at the roundtable, hosted by Future Processing, explored current challenges and future opportunities in the insurtech landscape.

During a discussion about ChatGPT, Piekos claimed that the tool would not be “prevalent” in insurance processes as it can only be “as good as the data you put into it”.

Opportunities?

ChatGPT is a chatbot developed by OpenAI that can answer questions in a conversational way.

The tool has been experimented with in the insurance industry, with insurtech Artificial Labs announcing earlier this year (31 May 2023) that it was using it as part of a pilot to assist underwriters.

And in March, insurer Zurich said it was experimenting with ChatGPT as it explored how it could use artificial intelligence (AI) technology for tasks such as extracting data for claims and modelling.

While Piekos highlighted that the tool can present challenges during experiments, he admitted there were also opportunities.

“There will be [possibilities] for companies to build a layer of specialised tooling on top of large language models to mitigate those deficiencies and probably enable a proper compliance layer on top of that,” he added.

However, he stressed that “on the other side of the spectrum”, customers needed to see “investment” in AI that can “support the customer journey in retail or commercial insurance”.