Is your voice worth… nothing? That’s the question David Greene, host of KCRW’s Left, Right & Center and former NPR personality, is asking as he sues Google over the alleged unauthorized use of his voice in the company’s NotebookLM AI tool. The narrative being pushed is about a radio host versus a tech giant, a David and Goliath story in the age of artificial intelligence. The real story here isn't about one man’s voice – it’s about the rapidly eroding concept of ownership in a world where AI can convincingly replicate anything.
Greene discovered the alleged imitation through a former colleague, finding himself “completely freaked out” by NotebookLM’s ability to generate podcasts in a voice eerily similar to his own. His January 23rd complaint in California doesn’t just allege a simple case of mimicry; it accuses Google of deliberately seeking to “replicate Mr. Greene's distinctive voice—a voice made iconic over decades of decorated radio and public commentary.” The suit hinges on violations of California’s right to publicity laws, which protect individuals from the unauthorized commercial use of their likeness, and unfair competition law. This isn’t about flattery; it’s about potential financial gain derived from someone else’s established brand.
Google’s response, delivered via spokesperson José Castañeda, is a flat denial. They claim the voice in NotebookLM is that of a “paid professional actor.” This is a crucial point, and one that feels… insufficient. Even if a voice actor was employed, the intent to create something that sounds like Greene – to leverage his established vocal persona – is where the legal and ethical lines blur. It’s not about the actor’s performance; it’s about the deliberate attempt to capitalize on a recognizable voice. The fact that Google isn’t addressing why they aimed for a soundalike is telling.
This article draws on reporting from mashable.com.
This case isn’t happening in a vacuum. Just last year, OpenAI was forced to pull its AI voice, “Sky,” after Scarlett Johansson rightfully protested its uncanny resemblance to her own. That incident, and the broader campaign against “AI slop and theft” launched by artists including Johansson in January, highlight a growing anxiety. We’re moving beyond concerns about AI replacing jobs to a more fundamental question: can you own your own identity in the digital age? The stakes are higher than many realize. It’s not just celebrities at risk; anyone with a distinctive voice, image, or writing style could find their persona replicated and monetized without their consent.
The legal precedent being set here is enormous. The Ziff Davis lawsuit against OpenAI in April 2025, alleging copyright infringement in AI training data, further complicates the landscape. These cases aren’t isolated incidents; they represent a systemic challenge to existing intellectual property laws. Current laws were designed for a world where copying required physical effort and distribution. AI changes everything, allowing for near-instantaneous and perfect replication at scale. The legal system is scrambling to catch up, and the outcomes will determine who controls the future of creative expression.
The implications for everyday users are significant. Imagine a future where AI-generated content floods the internet, indistinguishable from human-created work. How will we verify authenticity? How will we trust what we hear and see? The erosion of trust, coupled with the potential for widespread impersonation, could have a chilling effect on online communication and commerce. This isn’t a hypothetical scenario; it’s a rapidly approaching reality.
Watch closely for how courts interpret “distinctive voice” and “unauthorized use.” If Greene wins, it will establish a powerful precedent for protecting individual identity in the age of AI. But if Google prevails, it will signal a green light for companies to freely replicate and exploit the voices – and ultimately, the identities – of anyone with a public presence. The next six months will determine whether your voice is truly your own, or simply data waiting to be copied.
