Courtesy of Google
In a decisive move that escalates big tech’s role in generative audio, Google has embedded its latest music model, Lyria 3, directly into the Gemini app, allowing users to transform text, images, or video prompts into fully produced 30-second tracks complete with vocals, lyrics, instrumentals, and AI-generated cover art.
Currently in beta and available to users 18+ across eight languages, the tool positions Google squarely in competition with AI music startups like Suno, which continues to face high-profile copyright lawsuits over alleged unlicensed training data. Google, by contrast, is emphasizing compliance. The company says Lyria 3 was developed with copyright and partner agreements in mind and includes safeguards to prevent direct stylistic mimicry of specific artists.
Every track generated through Gemini is embedded with SynthID, Google DeepMind’s imperceptible watermarking technology, enabling later detection of AI-generated audio. Gemini users can also upload tracks to verify whether they were created using Google AI tools, an early signal that origin and authentication are becoming core platform features, not afterthoughts.
Strategically, the bigger play may be distribution. Lyria 3 is also powering enhancements to YouTube Dream Track for Shorts creators, giving Google an end-to-end ecosystem: creation (Gemini), distribution (YouTube), and verification (SynthID). The integration could shift leverage in licensing negotiations with rightsholders and PROs as generative audio scales.
For the industry, this isn’t just another AI launch; it’s a market signal. When a platform with YouTube’s global footprint operationalizes music generation at scale, the debate moves from “if” to “how revenue, rights, and royalties will be structured.”
About the Author
