The insurance industry is experiencing a seismic shift with the advent of artificial intelligence (AI) technologies. While AI promises unparalleled advancements in risk modeling and underwriting, it also brings forth a host of regulatory challenges, particularly as it relates to the use of third-party data and tools from vendors. Newly proposed AI rules suggest a new era of accountability and transparency for insurers is afoot, with the implication being that insurance companies will be held responsible for the information and data sets they buy from third-party sources.
At ZestyAI, we recognize the pivotal role of robust governance and compliance in facilitating the successful adoption of AI within the insurance sector. We identified the importance of regulatory approval early on, understanding its essential role in enabling our customers to leverage ZestyAI's products effectively. This realization fueled our commitment to not only understand but actively participate in shaping regulatory policy, laying the groundwork for a partnership approach to regulatory compliance.
This realization fueled our commitment to not only understand but actively participate in shaping regulatory policy, laying the groundwork for a partnership approach to regulatory compliance.
ZestyAI’s inaugural AI model, Z-FIRE, underwent a rigorous two-year approval process from rigorous regulatory bodies like the California Department of Insurance (CDI). This process involved comprehensive independent actuarial validations and bias testing, culminating in Z-FIRE becoming the first AI model ever approved by the CDI for wildfire rating. This achievement underscored the importance of our focus on regulatory leadership, prompting the establishment of a dedicated Regulatory function within ZestyAI. Since that initial approval, we have continued to successfully navigate regulatory approvals in multiple states, and Z-FIRE has now been approved for use in rating and/or underwriting in every state in the Western US.
As we anticipate the wave of AI insurance regulation across geographies, we are proactively preparing ourselves and our customers for compliance.
Here are some key elements of our strategy:
1. Extensive Model Validation
We conduct regular back-testing and independent model audits to ensure the continued robustness and reliability of our AI solutions.
2. Bias Testing
Addressing the crucial requirement highlighted in regulatory bulletins, we meticulously test for the absence of socioeconomic bias in our AI models, and avoid using any socioeconomic input variables, exemplified by our CDI-approved Z-FIRE model.
3. Scientific Data Selection
We collaborate with leading scientific and engineering organizations to ensure that our input data selection is rooted in hard science, mitigating the risk of bias.
4. Human-in-the-Loop Approach
Our "human-in-the-loop" approach ensures a balanced decision-making process between experienced underwriters and AI algorithms, enhancing transparency and accountability.
5. Engagement with Regulators
We maintain active engagement with key regulatory bodies, fostering transparency and facilitating adherence with evolving regulations.
6. Thought Leadership
Through publications and industry engagements, we strive to maintain a thought leadership position, advocating for regulatory frameworks that adapt to the new age of AI-driven risk modeling.
At ZestyAI, we firmly believe that AI-driven models hold the potential to revolutionize insurance underwriting and rating. Our proven track record of regulatory approvals, robust governance, and thought leadership position us to navigate future regulatory requirements effectively and drive enduring regulatory value in an ever-evolving landscape.
As the insurance industry embraces the era of AI regulation, ZestyAI stands ready to not only adhere to these emerging requirements but also proactively shape them, ushering in a new standard of transparency and accountability in AI-driven insurance.