Incident Overview
On January 23, 2026, veteran NPR presenter David Greene filed a lawsuit against Google LLC and its parent company Alphabet in Santa Clara County, California. The lawsuit alleges that Google’s NotebookLM tool, specifically its Audio Overviews feature, uses a synthetic male voice that mimics Greene’s vocal cadence and delivery without his consent or compensation. This marks the first known legal action under right-of-publicity laws specifically targeting an AI-generated podcast-host voice.
Greene’s complaint references an independent voice-recognition analysis, which reportedly found a 53–60% confidence level that his voice was used to train the AI model underlying NotebookLM’s audio summaries. The case draws parallels to a prior dispute involving Scarlett Johansson, who alleged that OpenAI’s Sky voice assistant replicated her voice without proper licensing.
NotebookLM’s Audio Overviews allows users to generate audio summaries of podcasts, including synthetic narration. Greene’s legal action suggests that the tool’s voice synthesis may have been trained on his voice without authorization, raising concerns about unauthorized use of vocal characteristics in AI training datasets. The lawsuit underscores broader ethical and legal questions surrounding the sourcing of training data for AI voice generation systems.
Google has publicly dismissed the allegations as baseless, asserting that the synthetic voice in NotebookLM is based on a paid actor rather than Greene’s voice. The company’s response contrasts with Greene’s claims, which emphasize decades of work perfecting his distinctive vocal delivery—a key aspect of his professional identity as an NPR host.
The incident highlights a growing tension between AI developers and content creators, particularly those whose voices or likenesses are valuable intellectual property. It also reflects a pattern of disputes over AI voice cloning, where synthetic replicas of human voices are deployed without explicit consent or compensation to the original voice owners.
Possible Root Cause
The core allegation in Greene’s lawsuit centers on the training data used to develop NotebookLM’s voice synthesis capabilities. Voice-recognition analysis cited in the complaint suggests that Greene’s vocal patterns may have inadvertently or intentionally been incorporated into the AI model.
This aligns with broader industry challenges in ensuring ethical data collection, where the absence of explicit consent or compensation for voice use can lead to legal and ethical conflicts.
This is particularly relevant in cases where the synthetic output is designed to sound indistinguishable from a human voice, as is the case with NotebookLM’s Audio Overviews.
If the voice in NotebookLM was indeed derived from Greene’s recordings, it suggests a failure in either data collection, annotation, or model training oversight. Alternatively, if the voice is unrelated to Greene’s, the lawsuit may stem from a misidentification or a deliberate attempt to replicate a familiar vocal style without proper attribution.
Harm and Legal Implications
Greene’s lawsuit alleges two primary harms: fraud and impersonation, as well as copyright and intellectual property infringement.
Fraud and impersonation arise from the potential misuse of Greene’s voice in a synthetic context without his consent. This aligns with right-of-publicity laws, which protect individuals from the unauthorized use of their name, image, or voice for commercial purposes. Greene’s case is notable as the first to apply these laws specifically to an AI-generated voice in a podcast context.
Copyright and intellectual property infringement are additional concerns. While Greene’s voice itself may not be copyrightable, the unique characteristics of his delivery—developed over decades—represent a form of intellectual property tied to his professional identity.
The incident also contributes to a broader pattern of disputes over AI voice cloning, where synthetic replicas of human voices are deployed without explicit consent. Greene’s lawsuit extends this pattern into the realm of podcasts and audio summaries, where the stakes for misattribution or unauthorized use are particularly high.
From a legal standpoint, the case tests the boundaries of right-of-publicity laws in the context of AI-generated content. The outcome could set a precedent for future disputes involving AI voice synthesis, particularly in cases where the synthetic output is designed to replicate human speech with high fidelity.
Takeaways for AI Developers and Policymakers
Google lawsuit offers several key insights for AI developers, particularly those working on voice synthesis and audio generation systems:
This is particularly important for high-profile individuals, such as Greene, whose vocal characteristics are distinctive and commercially valuable.
Third, the lawsuit serves as a reminder of the legal risks associated with AI voice synthesis. Even if the synthetic voice is not an exact replica, close similarities to a human voice can lead to claims of impersonation or unauthorized use.
This could involve updating right-of-publicity statutes, copyright laws, or other intellectual property frameworks to account for the use of AI in media production.
Google lawsuit is a pivotal moment in the evolving landscape of AI ethics and intellectual property.
Sources
Incident summary from the AI Incident Database, used under CC BY-SA 4.0.