Did Congress just make the first AI policy for a national legislature?

Last Thursday, the United States House of Representatives announced that it had adopted a police for the use of artificial intelligence, as part of the sixth Congressional Hackathon.

The current list of permissible AI use cases in the lower House of Congress isn’t public yet, but the Committee on House Administration (CAO) did release examples in a statement to Fedscoop, noting that they take a “risk-based” approach:

”…the generally permissible use cases are managing scheduling requests, translation, and transcription services. We cannot share any information related to security and privacy.”

“The CAO will not be recommending approval of AI for sensitive data or processes for the foreseeable future. However, we will learn from lower risk data sets as the technology evolves and matures.”

My former colleague Daniel Schuman says it might be the first AI policy of any national legislature anywhere — and he’d know.