Sen. Jamal Bennett of Illinois brought a worker-rights focus to Senate debate over artificial intelligence, arguing that consent cannot be considered meaningful when it is hidden in fine print or obtained through pressure.

Bennett’s remarks centered on the growing question of what employers may do with the work their employees produce as artificial intelligence systems become more deeply embedded in business operations. He said that writing, voice, design work and other attributable material should not be transferred for AI training without a clear choice for the employee beforehand.

The Illinois Democrat aligned himself with a central point raised by the preceding speaker: a purported agreement does not amount to genuine consent when workers are not plainly informed about what is being asked of them or feel unable to refuse.

That distinction is increasingly important as AI systems can be trained using material that reflects an employee’s labor, skill, identity and creative contribution. Bennett’s argument placed the responsibility on employers and institutions to make disclosure and permission real at the point a transfer is contemplated, rather than relying on terms that workers may not reasonably understand until after their work has been used.

The senator’s intervention reflects a broader Democratic emphasis on ensuring that technological change expands opportunity without stripping people of basic agency in the workplace. AI may be a powerful tool, but Bennett’s position is that innovation does not erase the need for workers to know when their contributions are being repurposed and to have a meaningful say before that happens.

His remarks also challenged a familiar corporate tendency to treat dense contractual language as a substitute for transparent decision-making. For workers, particularly those whose voices, writing or creative output may be directly traceable to them, the stakes can extend beyond an ordinary workplace policy. The question is whether people retain control over how the fruits of their labor are used to build new systems.

Bennett did not frame consent as an obstacle to technological development. Instead, he argued that a clear choice should be part of responsible development from the outset. That approach recognizes that confidence in AI will depend not only on technical capability, but also on whether the people whose work helps train these systems are treated fairly.

The Senate debate remains ongoing. Bennett’s comments put a straightforward principle before the chamber: workers should not be asked to discover after the fact that their identifiable work has been transferred into an artificial intelligence system. Consent, in his view, must come before the transfer—and it must be real.