Generative AI: an intellectual breakthrough, a security risk, or both?
OpenAI’s ChatGPT has become something of a poster boy for the explosion in impressive artificial intelligence (AI) products in the past few months. Generative AI, as the technology is known, promises huge commercial advantages and can already mimic human communications so well they can be virtually undetectable.
Of course, the re/insurance industry has been using various forms of AI for some time, although this has been based on models designed for a specific purpose, in a defined context.
But by making the prototype ChatGPT publicly available last November, its creators sparked imaginations, driving huge interest in its possibilities. The clamour was so great that competitor firms raced to get their generative AIs to market, with Google’s Bard and Microsoft’s Bing AI arriving mere months later.
Given its potential to accelerate claims and pricing models processes, to cut costs and increase revenue and to transform industry ecosystems, surely the industry should be embracing more AI. Or is this latest cycle of AI development just hype that will fizzle out?
Intelligent Insurer asked the experts what they thought generative AI could bring to the re/insurance industry table and about their concerns. This is what the experts, and ChatGPT itself, said.
A ‘breakthrough’ for unstructured data
ChatGPT demonstrates what is now possible with natural language AI (NL AI), representing “a breakthrough” for organisations that run on unstructured data, according to Charles Blanchet, managing director of Natural Language Solution, an AI integration consultancy.
His view, as you might expect, is that NL AI has widespread uses in the private and public sectors. “It is a platform technology that will develop into a vast ecosystem of data, software, and services,” he says, adding that it is “close to achieving ‘general intelligence’”.
When this happens, Blanchet says, it will have a larger impact on society than the internet.
“NL AI will transform all aspects of the insurance ecosystem, from the largest reinsurers to small retail brokers.
“Use-cases will focus on processes that consume and produce unstructured data. Brokers depend more on unstructured data than the other areas of the insurance ecosystem and thus they represent the largest transformation opportunity.”
