The Rise of AI in Insurance: Navigating Emerging Issues
Collegiate basketball teams captivated audiences this March, but another competition is stirring the property/casualty insurance world—a final four not of athletes, but of emerging issues. Verisk, a global data analytics leader, recently spotlighted these critical concerns during a webinar focused on the evolving landscape of AI in insurance. Artificial Intelligence was a standout topic, emphasizing its growing significance as state legislatures scramble to adapt.
AI Proliferation Prompts Legislation
The rapid advancements in AI technologies have triggered a wave of legislative action across the United States. In 2023 alone, numerous states introduced various forms of AI regulations, though only a select few have successfully enacted them into law. Laura Panesso, an associate vice president of government relations at Verisk, remarked during the webinar about policymakers responding swiftly to the challenges posed by AI.
“The trend reflects a burgeoning desire to regulate technology evolving at a pace we can barely keep up with,” Panesso noted. The National Association of Insurance Commissioners (NAIC) has also weighed in, issuing a bulletin that outlines expectations for the insurance industry regarding AI governance. This includes measures to identify biases and ensure fairness in decision-making processes. Notably, California, Colorado, and New York have begun to forge their own distinct pathways for state-specific AI regulations.
The State of AI Regulation: A Patchwork Approach
Without a cohesive federal framework for AI oversight, states are independently navigating the regulatory landscape. As of now, around 40 states have introduced or enacted some form of AI legislation, ranging from exploring its impacts to regulating specific uses. Major themes emerging in this burgeoning regulatory framework include requirements for AI deployers, ownership rights of data utilized for training models, and concerns about algorithmic pricing and discrimination.
One noteworthy development is Utah Senate Bill 26. This newly enacted law mandates transparency when generative AI is involved in consumer interactions, requiring disclosures and stipulating liability for consumer protection violations. Such measures underscore the urgency of addressing potential risks while acknowledging the incredible capabilities that AI technologies offer.
Edge Cases and Gen AI Hallucinations
While the potential of AI is substantial, the technology is not without its pitfalls. Greg Scoblete from Verisk highlighted concerns about “edge cases” and AI hallucinations. Edge cases occur when AI encounters scenarios for which it lacks sufficient training data. For instance, autonomous vehicles increasingly rely on AI systems for safety. Yet reports indicate these systems sometimes react dangerously when faced with unfamiliar situations.
In a troubling example from the UK, a luxury car with adaptive cruise control unexpectedly accelerated past a speed limit sign it didn’t recognize, leading to dangerous consequences. Scoblete emphasized that, unlike humans, AI lacks the instinctual understanding to navigate such anomalies safely.
Moreover, generative AI can produce errant outputs—referred to as “hallucinations.” These inaccuracies have already led to legal challenges, with over a hundred cases involving erroneous legal briefs generated by AI. The implications of inaccuracies in AI output raise critical questions about the norms of professional accountability and accuracy across various sectors, including law.
Legal Challenges and Emerging Liabilities
As AI continues to embed itself into more physical products, such as vehicles, the question of liability becomes increasingly complex. Scoblete pointed to emerging legal frameworks that may apply to virtual AI products, suggesting that existing product liability laws may need to adapt to encompass the unique challenges posed by AI.
He raised an important consideration: as AI technology becomes more integrated into our daily lives, its potential to cause property damage and personal injury grows. In a landscape where the regulatory framework is still developing, optimistically navigating these uncharted waters requires vigilance and adaptability.
The Road Ahead
While AI holds great promise in transforming industries, including property and casualty insurance, it brings with it a host of complexities that must be addressed. The current legislative efforts vary widely across states, and the industry must be prepared for ongoing discussions about the ethical deployment of AI technologies. As stakeholders engage in this evolving dialogue, the balance between innovation and regulation will remain a focal point.
As professionals in the insurance industry ponder the implications of these emerging issues, it’s clear that the intersection of AI and regulation will be a crucial area to watch in the coming years. The journey toward effective governance of AI technologies promises to be as intricate as the technologies themselves.