In August of 2020, the NAIC adopted five guiding principles for artificial intelligence (AI), based on the Organization for Economic Cooperation and Development’s AI Principles which have been adopted by 42 countries across the globe. These five principles include: Fair and Ethical Accountability, Compliance, transparency, security, safety, and robustness of outputs.
At the inaugural AAIS Pulse, AAIS VP Governance and Legal Compliance Robin Westcott was joined by Locke Lorde Partner Brian Casey for a preliminary discussion on what can be expected as regulators struggle to be able to regulate insurance contracts specifically around unfair discriminatory practices, one of the five principles.
Mr. Casey sees translating the regulation of AI into practical, measurable, regulatory practices as a tall order. “We’ll get there,” he says, but it’ll “take a few years”. In regard to the likely avenues the regulatory environment will use to monitor compliance with the new AI compliance, Mr. Casey predicts there will be examinations after the fact that look at conduct after the fact. This “presupposes”, he says that “the regulators know what they’re looking for and understand a little bit of what’s under the hood, so to speak.” Mr. Casey cites another method of potential monitoring as a sort of sort of self-compliance that would work until regulators have a better understanding of what they are and are not looking for.
Ms. Westcott believes we are at a “crossroads” around regulation and technology with “the work being done by the big data working group and the Information technology working group. Regulators struggle with how to get their hands around this newer technology and newer data elements that are coming in.” From her perspective, machine learning and AI will likely be no different. She predicts that regulators will struggle to have the “desired transparency that’s needed to look deeper beyond the technology to see the application.”
When asked how difficult he thinks it’ll be to get transparency around this type of technology, and to implement it in organizations, Mr. Casey believes it to be a double-edged sword. On one hand, “this is one of the correct guiding principles in this area and many others, privacy included.” On the other hand, “If it’s my technology and I’ve spent millions or billions of dollars developing the tech and enabling my workforce…I want to protect my intellectual property and the value I’ve developed,” he said. Overall, he predicts everyone will be able to weigh in on it and will eventually have to come to terms with the decision. Some level of transparency is necessary, according to Mr. Casey.
Regulators are up for the challenge, Mr. Casey ensured, as he described insurers as hardworking and emphasized their enjoyment of learning. The industry must work together however, avoiding ‘silo-ing’ themselves and dampening their full capabilities.