Stability AI on September 13 launched a new AI product called Stable Audio that can generate audio from text prompts. For example, we used the model to generate the following audio using the prompt “calm meditation music to play in a spa lobby.”
In the free version, users can generate 45-second tracks up to 20 tracks a month and with the $12/month Pro version, users can generate 90-second tracks up to 500 tracks a month.
Stability AI is popular for its widely used text-to-image generator Stable Diffusion, but the company faces a lawsuit filed by a group of artists for copyright infringement. The artists allege that Stable Diffusion’s AI models were trained on billions of copyrighted images without compensation or consent from the artists. “If Stable Diffusion and similar products are allowed to continue to operate as they do now, the foreseeable result is they will replace the very artists whose stolen works power these AI products with whom they are competing,” the complaints stated. Stability AI has, however, filed its own motion to dismiss the suit.
The company’s latest AI audio generator might also be subject to similar copyright lawsuits given that it’s trained on audio created by someone else that is possibly copyrighted. Stable Audio is not the first of its kind though. Google also launched a text-to-audio generator called MusicLM in February. Back then MediaNama spoke to copyright lawyers to understand if the works produced by AI generators infringe on copyright. You can read that story here.
- Six Key Arguments By Stability AI Over Allegations Of Copyright Violation By Artists In US
- MusicLM: Five Key Points By Copyright Lawyers On Google’s AI-Based Music Generator
- Why Is Universal Music Group Asking Streaming Services To Block AI-Generated Music?
- US Copyright Office Seeks Public Views On Study Into AI And Copyright