Regulating artificial intelligence isn't just prudent—it's urgent

AI expert Chloe Autio explains how to develop AI policy that's practical, enforceable, and flexible.

The conversation around AI regulation is predominantly polarized, often focusing on catastrophic outcomes or overly optimistic possibilities. However, the reality lies somewhere in between. As such, it’s crucial to understand and classify these risks and gauge how they can be addressed practically and effectively, considering both the present applications and future implications of AI technologies.

“We know how to manage [emerging technology]. We know what to do. Let's make a plan and do it. And put resources and action toward it,” Chloe Autio, director of policy at The Cantellus Group, told me in a recent interview. “This is what I do in my daily job, helping different types of organizations, customers, and clients, understand the applied technology use in front of them, what problem they're trying to solve with that technology, and the risks that can result from those applications."

Regulations must ensure a balance of encouraging innovation while protecting privacy, fairness, and democratic norms. Congress has proposed multiple frameworks, and advisory committees have offered various recommendations, but action has been slow. It's time we move beyond discussions and into implementations, said Autio. 

Applying existing laws to new AI contexts, such as anti-discrimination laws and unfair practices, could be a good starting point. A fundamental element missing from current regulatory conversations is the establishment of standard practices and industry norms. Regulatory guidance on practical aspects, such as what an AI impact assessment should look like or who should be in charge of AI initiatives, could prove invaluable in aligning AI technology with our societal needs.

We are teetering on the edge of an AI-powered economic revolution. Autio says it's critical to have robust regulatory frameworks to guide this powerful technology's development and application, mitigating its risks and harnessing its vast potential.