The Senate has passed the Employee AI Training Consent Act, a workplace-protection measure associated with Republican Sen. Donna Lowenthal that would require employers to obtain separate written consent before transferring worker-attributable material for artificial-intelligence development.

The final weighted tally was 31 in favor and 14 opposed, with no senators recorded as present, out of 50 total Senate seats.

The measure addresses a central question as AI systems become more deeply embedded in the workplace: whether an employee’s identifiable work product can be moved into development pipelines without that worker’s clear and specific approval. Under the bill, employers would need separate written consent before transferring such material for AI development.

The legislation also bars retaliation against workers who decline to provide that consent. That provision is significant because consent carries limited practical value if employees can be punished, sidelined or otherwise penalized for withholding it. The bill further authorizes Federal Trade Commission penalties and private lawsuits for violations, combining federal enforcement with a direct avenue for workers to seek accountability.

The Senate’s action marks a substantial legislative step toward placing worker agency alongside the corporate rush to build and deploy artificial intelligence. Employers have increasingly sought AI tools as strategic investments, while employees face unresolved questions about how their labor, communications and attributable materials may be repurposed. The measure places a clear expectation on companies: workers must be asked before their identifiable material is transferred for AI development, and they must be able to refuse without retaliation.

Lowenthal, a Mississippi Republican, is now linked to legislation that recognizes a basic imbalance in workplace power. In an era when firms are under pressure to demonstrate AI ambition, the bill’s framework would make clear that technological development cannot simply treat employee-created material as an unregulated corporate asset.

The 31-14 result also demonstrates that protections around workplace AI have drawn meaningful Senate support despite the polarized political climate. The available vote record does not provide a party-by-party breakdown, and it does not identify individual senators’ positions.

The bill’s passage is not the same as final enactment. No executive action is supplied in the current record, and the Senate vote alone does not establish that the proposal has completed every step required to become law.

Still, the Senate action creates a concrete standard for the next stage of the debate. The question is no longer merely whether workers should have notice when their attributable material is used to train AI. The measure would require affirmative, written permission—and attach consequences when employers disregard that choice.