OpenAI has reportedly been taking steps to develop an AI that generates music

OpenAI, a leading artificial intelligence research and deployment company, is developing an advanced AI system designed to generate music from text or audio prompts. This initiative places OpenAI in direct competition with rising players in the AI music generation sector, including Suno and Udio, marking a major new step in the broader contest for AI innovation. The forthcoming model is reportedly capable of producing a range of musical outputs — from soundtracks to full accompaniments — signaling an expansion of OpenAI’s ecosystem beyond its established text and image generation platforms. This move into audio generation follows the trajectory set by its text-to-video model, Sora, and reflects the company’s ongoing effort to diversify its suite of generative AI tools.

OpenAI’s exploration of AI-driven music is not unprecedented. In 2019, the company introduced MuseNet, a deep neural network capable of composing four-minute musical pieces with up to ten instruments, blending genres from classical to pop and rock. The following year, it released Jukebox, a more advanced neural network that generated raw audio tracks featuring basic singing. Jukebox was trained on a dataset of approximately 1.2 million songs and could produce music across various genres and artist styles based on user inputs such as genre, artist, and lyrics. Despite these innovations, both MuseNet and Jukebox were primarily experimental and faced notable constraints in sound fidelity and structural coherence, limiting their use in mainstream products like ChatGPT.

Today’s AI music landscape is far more mature and competitive than during OpenAI’s earlier experiments. Companies such as Suno and Udio have attracted widespread attention for generating high-quality, full-length songs from simple text prompts. Their platforms enable users with no musical background to create complete songs — including vocals and instrumentation — democratizing access to music production. Major technology firms, including Google, are also pursuing their own music generation systems, such as Lyria, underscoring the growing commercial and creative potential of this field. The rapid pace of development reflects expanding opportunities for AI-generated music, spanning applications in entertainment, advertising, and digital content creation.

However, the rapid growth of AI music generation has intensified legal and ethical debates surrounding copyright and intellectual property. These systems rely on vast datasets of existing music, raising concerns about the use of copyrighted material without authorization. Major record labels — including Sony Music Entertainment, Universal Music Group, and Warner Records — have filed lawsuits against both Suno and Udio, alleging that their models were trained on copyrighted songs without permission. These legal challenges underscore the broader conflict between the data demands of large-scale AI systems and the rights of artists and rights holders. The outcomes of these cases could shape the regulatory and commercial environment for generative AI, potentially constraining data access and influencing how future AI music platforms operate. Adding further complexity, U.S. copyright law currently does not recognize works produced entirely by AI as eligible for protection, leaving unresolved questions about ownership and monetization in the emerging era of AI-generated music.