The Senate approved the Employee AI Training Consent Act, advancing a federal proposal that would require employers to obtain separate written consent before transferring worker-attributable material for artificial intelligence development.

The measure, sponsored by Donna Lowenthal, passed 31-14. There were no senators recorded as present, and the bill now moves to the President’s desk.

The legislation places workplace consent at the center of a growing policy dispute over how employers may use material tied to their workers as AI development expands. Under the proposal, employers could not transfer such material for AI development without first obtaining written consent that is separate from other workplace agreements.

The bill also would prohibit retaliation against workers who refuse to provide that consent. That provision would make a worker’s decision to decline participation a protected choice under the proposed framework, rather than simply a matter for an employer’s internal policy.

In addition, the measure authorizes penalties through the Federal Trade Commission. The FTC role gives the proposal a federal enforcement component, moving it beyond a disclosure or voluntary-compliance model.

The Senate vote is a significant legislative step for an issue that has increasingly tested the boundaries between employer control of workplace systems and employees’ interests in material connected to their work. Artificial intelligence policy has often involved broad questions about innovation, privacy and accountability; Lowenthal’s bill focuses more narrowly on the conditions under which worker-attributable material can enter AI development.

The final tally showed support sufficient to clear the chamber despite opposition from 14 senators. With the vote complete, the central question is no longer whether the Senate will advance the consent requirement, but whether the President will sign the legislation, allowing its written-consent, anti-retaliation and FTC-penalty provisions to take effect.

The bill does not broadly prohibit AI development or employer use of technology. Instead, it would establish a defined consent requirement before employers transfer material attributable to workers for development purposes. That distinction could be important as companies and employees confront more direct questions about whether workplace activity, output or other attributable material may be used to train or develop AI systems.

For Lowenthal, the Senate’s action puts a worker-protection measure on the President’s desk at a time when AI policy remains unsettled across much of the federal government. The bill’s structure pairs a consent mandate with a retaliation ban and potential federal penalties, creating a more formal set of obligations for employers than voluntary workplace standards alone.

The vote also illustrates how AI legislation can draw attention outside the most familiar debates over consumer privacy or the technology sector. By centering workers, the measure frames AI development as a workplace issue as well as a technology-policy question.

The President now has the measure for consideration.