OpenAI is reportedly developing a new generative-music tool that could allow users to produce original songs or scores using both text and audio prompts. According to multiple sources, the project is designed to let creators type in a description like “jazzy piano ballad with soft strings” or upload a vocal track and ask for a backing arrangement; the system would then generate a finished piece of music tailored to those cues.
The new tool represents a strategic shift for OpenAI, which has previously explored music-generation models (such as Jukebox) but has largely focused until now on text, image and audio-speech models. With this initiative, the company appears to be targeting the creative-media market more aggressively—an area where generative AI is already gaining traction and scrutiny from artists, rights holders and regulators alike. Training for the system is said to draw on annotated scores and musical-structure data, and insiders suggest that OpenAI has collaborated with music-school students and professionals to label and organise training sets.
If launched, the music tool could have multiple uses: video producers might generate soundtracks, vocalists might get instant accompaniment, or hobbyists could create original tracks without formal musical training. However, the project enters a contentious space: the music-industry has raised serious concerns about AI-generated content, particularly around copyright, attribution, licensing and the impact on songwriters and performers. OpenAI will need to address these issues—whether through licensing deals, transparency about training data, or controls on how generated music is used—before widespread release.
It remains unclear when the tool will be made publicly available or whether it will be offered as a standalone product or integrated into existing OpenAI services like ChatGPT or its media-generation suite. The development does, however, reflect broader trends: as generative AI matures, companies are moving beyond novelty into tools that sit at the core of creative workflows. For OpenAI, success may depend not only on the quality of the output but also on how it navigates the creative-rights ecosystem while delivering something genuinely useful for musicians, creators and media professionals.
