David Greene, the former host of NPR’s “Morning Edition,” claimed in a lawsuit that Google stole his voice to train its AI audio tool to produce audio clips using his voice.
Greene told the Washington Post in a story published Sunday that he was alerted to Google’s NotebookLM tool when he started receiving messages from friends asking if he licensed his voice to the company. The tool generates podcast-like audio summaries with distinct male and female voices, and it was used by Spotify last year to produce podcast episodes related to users’ “Spotify Wrapped” summaries.
“I was, like, completely freaked out,” Greene told the Post, pointing to verbal tics he claimed the voice emulated. “It’s this eerie moment where you feel like you’re listening to yourself.”
Greene filed the lawsuit last month in Santa Clara County, California, alleging the company trained the tool using his voice without compensating him. He cited an analysis by an AI forensic firm that offered a confidence rating between 53 percent and 60 percent that his voice was used. (A rating above zero means the voices are similar, according to the Post.)
Aside from missing opportunities to capitalize on his voice, Greene said part of his motivation in filing the suit was the surreal experience of listening to his voice without his participation.
“My voice is, like, the most important part of who I am,” he said.
A Google spokesperson told TheWrap that the allegations were “baseless” and that the company hired an actor to train the tool.
Similar cases have come up throughout the last half-decade as AI companies experiment with tools that mimic human conversations. Scarlett Johansson tore into OpenAI in 2024 after she said its virtual assistant sounded similar to her voice — despite repeatedly rejecting the company’s requests to license it. Le Creuset also denied its participation with an online ad campaign that used an AI-generated Taylor Swift to sell cookware.
