Meta Platforms, the parent company of Facebook, has introduced an open-source AI tool known as AudioCraft. This innovative tool is designed to empower users to generate music and audio content based on text prompts.
Comprising three distinct models: AudioGen, EnCodec, and MusicGen, AudioCraft is equipped to cater to a variety of functions, including music creation, sound generation, compression, and more. The MusicGen model, for instance, has been trained using a combination of proprietary Meta-owned music and licensed tracks.
Meta’s move into the AI-powered audio creation arena is not without its challenges. Concerns have arisen within the music industry about potential copyright violations, a common issue with machine-learning software that relies on patterns extracted from web data. Critics argue that such AI systems could inadvertently replicate copyrighted content without proper authorization.
This development comes in the wake of Alphabet Inc’s foray into AI-generated audio with its experimental tool called MusicLM. As technology giants like Meta and Alphabet continue to push the boundaries of AI-powered creativity, questions around intellectual property rights and the ethical use of AI-generated content are likely to remain at the forefront of discussions within both the tech and creative sectors.