Articles

A top Democrat is calling on the Trump administration to halt the implementation of a new artificial intelligence (AI) system designed to streamline air travel.

Rep. Don Beyer (D-Va.) warned that the Federal Aviation Administration’s (FAA) new AI tool, which aims to eliminate flight delays and cancellations, is being launched without assurances of public safety.

“The FAA should not use my constituents as guinea pigs for an unproven AI air traffic control system,” Beyer said Monday in a statement. “The safety of the 24 million people traveling through DCA each year must be the top priority.”

The warning echoes the broader concerns on Capitol Hill surrounding the development and implementation of AI, particularly after several leading industry voices sounded alarms this month about the potentially existential dangers the rapidly growing technology poses to humans.

The new FAA tool, dubbed the Strategic Management of Airspace Routing Trajectories (“SMART”) system, is designed to crunch a host of factors — including weather patterns, airport capacity and real-time airspace conditions — to help manage air traffic and ease congestion at the nation’s airports.

The system launched on Monday at three airports around the nation’s capital — Reagan Washington National (DCA), Dulles International (IAD) and Baltimore/Washington International Thurgood Marshall (BWI) — with future plans to expand nationwide.

“By fundamentally reshaping how we manage our airspace and preventing problems before they happen, SMART will slash those frustrating delays, reduce stress on air traffic controllers, and lower travel prices,” Transportation Secretary Sean Duffy wrote Monday in a statement.

Beyer is skeptical. The six-term House veteran is a co-chair of the Congressional AI Caucus, a bipartisan group that seeks to advise fellow policymakers on the emerging technology. And he also represents the Northern Virginia district that’s home to DCA.

Beyer is urging the FAA to suspend the rollout of the tool until it can be vetted more thoroughly by human operators on the ground — “at the very least until it is clearly established beyond doubt via air traffic controller feedback and independently verified stress testing that the technology is safe.”

“Dangerously rushing a system that holds Americans’ lives in its hands is not a good use case for artificial intelligence,” he said.