NAIC approves model bulletin for insurer AI use
By onAnnouncements
As technology deployment increases to transform all stages of insurance interaction with consumers, the National Association of Insurance Commissioners (NAIC) recently approved a model bulletin on the use of artificial intelligence (AI) by insurance companies, an NAIC press release states.
The “Model Bulletin on the Use of Artificial Intelligence Systems by Insurers” passed during the association’s 2023 Fall National Meeting held earlier this month. It outlines the need for processes and controls to prevent possible AI inaccuracies, discriminating biases, and data vulnerabilities.
“This initiative represents a collaborative effort to set clear expectations for state departments of insurance regarding the utilization of AI by insurance companies, balancing the potential for innovation with the imperative to address unique risks,” said Maryland Insurance Commissioner Kathleen Birrane, who is the chair of NAIC’s Innovation, Cybersecurity and Technology Committee and drafted the bulletin.
The bulletin reminds insurers of established regulatory laws, such as the Unfair Trade Practices Model Act, that regulate unfair methods of competition or unfair or deceptive acts. It states governance and controls on AI systems are needed to comply with these laws.
“Compliance with these standards is required regardless of the tools and methods insurers use to make such decisions,” the bulletin says.
It also informs insurers about the steps state insurance departments could take to ensure that the companies follow the law. This could include documentation that insurers have set up governance and risk management frameworks for AI systems.
In an interview with Repairer Driven News, Birrane said the bulletin gives insurers guidelines on following the law while using AI. The bulletin does not mandate specific procedures or controls that insurance companies must use. This allows for the companies to be flexible depending on their company’s size and needs, she said.
“We think it is good that AI can develop and provide products,” Birrane said Tuesday. “They help develop and provide pricing that makes sense. They allow companies to access data quickly. They can write in areas that are challenging.”
Birrane says private insurers used to avoid issuing flood insurance because it was viewed as too big of a risk.
“But now, using much larger data to identify trends, you have an emerging private flood insurance market,” Birrane said. “Mining that data to predict outcomes is valuable and important, but we also know these methodologies can have broad ramifications that are harder to understand.”
Car insurance also has changed as it has become easier to analyze data, Birrane said. Historically, companies used fewer data sets to predict the riskiness of a driver. For example, insurers might review a driver’s age, driving history, and years of driving experience.
“In the simple world, you might have five or 10 rating factors,” Birrane said. “Now, they might consider 150 things about me.”
Car insurers have expanded the way they use data to include algorithms that search for potentially exaggerated claims, Birrane said; echoed by the NAIC press release elaborating that, “AI techniques are deployed across all stages of the insurance life cycle, including product development, marketing, sales and distribution, underwriting and pricing, policy servicing, claim management, and fraud detection.”
Insurers are already using AI-based photo analysis. CCC Intelligent Solutions, as an example, recently launched new AI tools for insurance providers that use photos of vehicle damage to generate predictions.
One product, First Look, uses AI to predict the point of impact or whether a vehicle is likely repairable. Another, called Impact Dynamics, uses AI to predict the change of velocity at impact, which can help an insurer evaluate potential injury, company officials have said.
Insurers shouldn’t use these tools to make claim decisions or determinations, the company said. They are tools that insurers can use as part of claim management processes.
Birrane said using AI during claim processes can be helpful if caution is used.
“It is not an improper methodology,” Birrane said. “As long as the company has a human being in the loop, that is. Machines can’t earn medical degrees or be medical directors.”
A recent study by YouGov, on behalf of Sprout.ai, showed consumers remain skeptical of AI use during the claims process, with 51% of 4,442 U.S. and UK consumers surveyed saying they were not likely, or very unlikely, to switch to an AI company offering a faster claims process.
The survey respondents said trust was the No. 1 priority when choosing an insurance provider.
The insurance sector is primarily regulated at the state level, Birrane said. The bulletin is one step in providing oversight to the changing market.
“It is important that we evolve and articulate U.S. guidance because of the rapid evolution of these tools,” Birrane said.
Birrane said 22 states participated in the drafting of the model bulletin. The process also included multiple public input opportunities.
The next step will be reviewing the statutory authority states can use to oversee AI systems, Birrane said.
Images
Featured image courtesy of Poca Wander Stock/iStock