A witness testifying at a Senate hearing said AI-generated songs that mimic real singers are appearing on music charts and posing a threat to working musicians.
The witness, who identified themself as representing and advocating for musicians, told the panel that "people charting in the top 10 are AI songs," citing the gospel chart as an example of tracks attributed to fabricated artists. The witness described the practice as taking musicians' voices and "making things that sound exactly like us, but put new words and new messages" into those performances.
The witness explained that an "AI song" often involves a fake identity or brand — "a fake person" — created by training software on other vocalists and songs, then producing a new recording that is marketed as belonging to an artist who does not actually exist. "Claim it's an artist," the witness said, adding that such tracks can sound good enough to attract large numbers of streams and reach chart positions. "That's deeply threatening," the witness said, emphasizing that "very little" has been done so far to address the issue.
During questioning, a member of the panel framed the problem concretely, saying, "you don't want words to be coming out of your mouth that you ever said," and asked the witness to explain recent reports of AI-created "world famous" or "bestselling" singers. The panelist later concluded the exchange by saying, "We're going to work with you in any way we can," indicating a willingness to follow up.
The hearing excerpt included no references to specific statutes, pending legislation, or formal votes. The witness urged further attention from lawmakers and industry stakeholders to address how AI is being used to replicate and rebrand performers without clear attribution or consent.
The hearing provided a firsthand account from the music community of a new use of generative AI and signaled possible next steps of committee engagement; no formal policy or regulatory action was recorded in the provided transcript excerpt.