Sean Parker steers Stability AI toward licensed music tools

Stability AI is refocusing its business on music-making tools for professional users, with Sean Parker and chief executive Prem Akkaraju seeking to make the company a leading supplier of generative AI for musicians. The shift pairs newly released audio models and editing software with financing from major entertainment companies and licensed music data. Parker, the Napster co-founder, joined an $80 million rescue of Stability AI in 2024 after the company nearly collapsed amid overspending and internal turmoil. Founder Emad Mostaque was ousted, and Akkaraju, a longtime friend of Parker’s, became chief executive. In comments reported by TechCrunch, Parker said he was “playing by the rules this time,” acknowledging that his earlier approach of seeking forgiveness rather than permission had not worked well. He said the objective was to make Stability the “go-to AI toolmaker for music professionals.” The company announced a $76 million Series B round on Aug. 25, bringing its total funding to $232 million across two equity rounds and convertible notes. Stability AI said the money would support its creative-production product suite, applied research and professional-services arm, as well as tools for music, gaming and entertainment. The investor group included Sony Music Group, Universal Music Group, Warner Music Group, Electronic Arts, AMD Ventures and Pacific Alliance Ventures. Parker, Eric Schmidt, Coatue, Greycroft and Kadmos Capital returned after investing in the preceding round; WPP had already been a strategic partner and investor. Coatue co-founder Thomas Laffont joined the board following the financing. The financing also deepens Stability AI’s ties to rights holders. Sony, Warner and Universal licensed music catalogs for training as part of the funding deal. Variety separately reported that the investment followed agreements with Universal, Warner and Electronic Arts to build AI models using their catalogs and intellectual property. Stability AI has released three new audio models and AI music-editing software. Its tools can generate complete instrumental tracks or shorter musical passages from text prompts. Stable Audio 3.0, introduced in May, is a family of open-weight music models trained on fully licensed data, according to the company. Artists can access Stable Audio 3.0 through a Digital Audio Workstation plug-in or StableAudio.com. Its features include changing individual or multiple audio segments and extending a recording beyond its original end point. The May announcement described four models, three with downloadable weights, including models for sound effects, on-device music generation and longer tracks. The company has also published instructions for users to customize its Small and Medium models with their own libraries through LoRA fine-tuning. Its Community License permits users to distribute and commercialize outputs within stated conditions, while organizations with more than $1 million in annual revenue need an enterprise license for commercial coverage; that license includes legal indemnification. The music focus marks a broader expansion from Stability AI’s 2022 release of Stable Diffusion, its open-source text-to-image model. The company has continued to emphasize open-source models, while presenting Stable Audio’s licensed-data approach as an alternative to AI music generators initially trained on unlicensed tracks. A planned update would let musicians direct generation using a hummed melody or a beatboxed drum pattern. The controls remain an announced feature rather than a release, and Stability AI has not set a launch date.