What if, 12 months ago, someone bet you that a technology was coming in the next year that would cause great excitement and great fear in all businesses, industries and geographies? Would you have taken that bet?
Of course, now we know that the technology is generative AI. Popularized by chatbots like ChatGPT, generative AI has left many, including insurers, grappling with its implications and seeking opportunities.
“Everybody is still in shock to an extent because nobody really realized that what is happening right now could have been even possible 12 months ago,” says Jose Ribeiro, Chairman of the Board of Starr Insurance Companies.
The opportunities and risks associated with generative AI have become a top issue for insurance board directors.
Ribeiro shared his thoughts with SAS about the impact of generative AI in insurance and what leaders should consider. Here are three key points he highlighted with us.
1. Generative AI is different
Unlike many technologies, generative AI knows no boundaries. It is global, spans industries and finds application in almost all business functions. Thus, its use must comply with both company guidelines and local regulatory rules. “The difference with generative AI technologies is that it will be used widely, by everyone around the world, across countries and businesses,” Ribeiro said. “Companies that are not prepared to have responses to the use of these technologies will be at a disadvantage.”
The insurance sector stands to gain from generative AI due to its inherent reliance on data-driven decisions, risk assessments and customer-centricity. However, generative AI introduces unbounded risks that insurers must consider if used.
2. A ‘double whammy’
For insurers, generative AI is a technology that will affect the operation of all functions. However, unlike other industries, insurers will need to assess its impact on their operational processes and evaluate its effects on the insurance product itself.
As Ribeiro emphasized, this requires a thorough risk assessment to anticipate and mitigate potential consequences. “In claims, underwriting, investments, across all functions of an insurance company, let’s do a risk assessment of what is going to happen,” he said.
For example, consider the recent situation where an attorney used ChatGPT to prepare legal briefs. Unfortunately, the cases cited were fictional. For most businesses, that wouldn’t matter. However, the insurer who sold that attorney's professional liability policy will likely have a claim. This highlights that unique challenge insurers face when it comes to generative AI. This risk of misinformation or inaccuracies could lead to legal disputes and potential claims against the insurer.
Ultimately, while generative AI can help improve insurers’ operations and products, boards must also address its ethical and regulatory standards. Doing so helps insurers proactively adapt and safeguard operations and maintain trust with policyholders.
3. What insurance boards should do next
Ribeiro shared that the boards he serves are learning about the technology as much as possible. He approaches generative AI from a risk management perspective. He wants to understand what the risks are to the business - to the integrity of its data, reputation and profitability. “As a board member, I want to understand this from a risk management perspective,” Ribeiro said. “Some uses of these technologies can be very, very good, but some can be very, very bad.” His goal is to develop a strategy that enables companies to use generative AI to remain competitive and to make it a better business.
Some prognosticators suggest that the impact of generative AI will be equal to that of the printing press or the internet. Its tremendous potential supports such an evaluation. Despite its eventual ranking on the list of innovations, the need for insurers to address it from a risk management perspective is immediate.