AI-generated audio is permitted on Freesound if tagged with GenAI and the generating model is named in the description

Freesound permits upload of audio created by generative AI models, subject to two disclosure requirements: the sound must carry the ‘GenAI’ tag, and the description must identify which AI model produced the audio. These requirements make AI-generated content transparent and filterable, allowing users who want to avoid AI-sourced sounds to do so. The standard CC license selection still applies — an uploader must choose CC0, CC-BY, or CC-BY-NC just as for any other sound. There are no additional restrictions beyond tagging and disclosure. This policy distinguishes Freesound from platforms that prohibit AI-generated content entirely, while still maintaining a transparency layer.

Examples

An uploader creates a rain ambience using a generative audio model, exports it, and uploads to Freesound. They tag it ‘rain ambience GenAI’, select CC0, and write in the description ‘Generated with AudioLDM2 v2’.

Assessment

A creator uses a text-to-audio AI model to produce 10 sound effects and wants to upload them to Freesound under CC-BY. What two things must they do that differ from a standard upload of recorded sounds?