In a significant legal development, the Colorado Department of Insurance has ushered in a new regulation centered on insurers’ use of artificial intelligence models. With this, Colorado has taken a lead in tackling race-based discrimination through AI. The rule is likely to be emulated by other states, given the nationwide repercussions of such advanced regulatory guidance.
The novel Colorado regulation, detailed more comprehensively here, addresses a critical concern: the intersection of technology, insurance, and discrimination. It broadly aims at regulating the use of predictive models by insurers and nudging the sector towards an era of transparency and inclusivity.
In essence, the rule positions Colorado as a powerful example for states grappling with similar questions around fairness in AI usage. Moreover, such a crucial regulatory development may invariably have implications on large corporations, which will need to adhere to them for their operations in Colorado. And, if the approach replicates beyond Colorado, they may need to adjust their practices on a broader scale.
Concerning the specifics of the regulation, it primarily prohibits insurers from employing an AI system or predictive model for underwriting or ratemaking if it violates Colorado’s Unfair Practices Act. Moreover, there’s a key emphasis on transparency, as insurers will need to disclose model names and versions, underlying theories, and key assumptions, among other critical metrics. Still, it remains to be seen how this regulatory shift will shape the insurers’ interaction with AI technology and its potential to mitigate or exacerbate discrimination.
Undoubtedly, the spotlight from Colorado’s new rule will spur discussions on AI, privacy, and discrimination at the national level. Based on the pace at which AI is transforming sectors, other states may soon follow suit, giving rise to a comprehensive, country-wide AI regulation.