Zoë Baird urges Congress to launch an AI risk commission

Zoë Baird, a senior fellow at the Center for American Progress and former senior counselor for technology to US Commerce Secretary Gina Raimondo, argues in Tech Policy Press that Congress should quickly establish a bipartisan public-private AI Commission and give it six months to deliver an initial report.

Her case rests on an unusual signal from industry itself. "Industries do not typically lobby for their own regulation or warn that their products could cause catastrophic harm," she writes. "The people building artificial intelligence are the ones sounding the loudest alarms. That is not normal."

Baird would have the commission answer four questions: the federal government's role and the capabilities it already has or urgently needs to build; the responsibilities of AI developers and those who deploy AI systems and the enforcement mechanisms that should attach to each; what "information-sharing architecture is needed — between developers and government, between government agencies, and between the US and its allies"; and which global alliances are most urgent, including "agreements with China as the other leading AI developer with its own reasons for avoiding unconstrained AI and the loss of human control."

She reaches for historical precedent: the Aspin-Brown Commission after the Cold War, the 9/11 Commission, and the 2002 Markle Foundation task force on National Security in the Information Age, which produced a blueprint shifting the intelligence community from a "need to know" to a "need to share" posture. Some steps, Baird adds, should not wait on a commission report. "Certain steps should be taken right away, like requiring embedded independent reviewers and government safety review."