AIVA policy evolution
Before/after stance changes across captured policy versions, with exact citations. If no before/after delta is available yet, AIRIN shows the latest citation-backed stance events instead.
No before/after stance delta is available for this filter yet. Latest citation-backed stance events are shown below.
Jun 17, 2026trainingmedium
content license
Latest stance: broad license
“Licensor provides Licensee with the ability to upload their own MIDI or Audio compositions (the “Uploaded Influence") to Licensor’s Servers, so that AIVA can generate a unique MIDI and Audio Composition that shares some of the emotional qualities of the Uploaded Influence (the “Upload Influence Feature”). Licensee provides Licensor with a worldwide, non-exclusive, commercial and transferrable license to train AIVA's AI systems on the Uploaded Influence, in perpetuity. Licensee warrants to Licensor that neither the Uploaded Influence, nor any portion thereof, shall infringe upon the rights of any third party. Licensee shall take full responsibility for its usage of the Uploaded Influence in the context of the Upload Influence Feature, and shall hold harmless Licensor against any third party claim.”
Open citationJun 17, 2026traininghigh
model training
Latest stance: training permitted
“Licensee is not permitted to use the MIDI or Audio compositions to set up a competing / similar service to what is provided by AIVA on the Site. Licensee is not permitted to use the music as part of a large scale upload or licensing to any third party. For example, large scale licensing of MIDI or Audio compositions that have not been modified or curated is not permitted. This use case is better served through an API integration, so please contact us if your use case requires a high volume of compositions & an automated pipeline. Licensee is not permitted to use the Audio and/or MIDI Composition, or any audio sample as part of a training dataset for any Machine Learning, Deep Learning or statistical algorithm. If the Licensee wishes to use the Audio and/or MIDI Composition as part of a training dataset, this use case would be ruled by a separate Licensing Agreement, to be negotiated and signed between the parties. Licensee is not permitted to use the Audio, MIDI Composition, Chord files, or any audio sample as part of a large scale database of music that is beyond the scope of this Agreement, and beyond the scope of individual use. Licensee is not permitted to bypass limitations imposed by each subscription plan in any way, including but not limited to by registering for an account using temporary email addresses or aliases. To ensure the best quality of service, and to ensure that the servers do not get overloaded, the Licensor has put certain limitations in place on the Services, which are as follows: A single user cannot create more than 5 Compositions at one time.”
Open citationJun 17, 2026traininghigh
model training
Latest stance: training permitted
“Licensor provides Licensee with the ability to upload their own MIDI or Audio compositions (the “Uploaded Influence") to Licensor’s Servers, so that AIVA can generate a unique MIDI and Audio Composition that shares some of the emotional qualities of the Uploaded Influence (the “Upload Influence Feature”). Licensee provides Licensor with a worldwide, non-exclusive, commercial and transferrable license to train AIVA's AI systems on the Uploaded Influence, in perpetuity. Licensee warrants to Licensor that neither the Uploaded Influence, nor any portion thereof, shall infringe upon the rights of any third party. Licensee shall take full responsibility for its usage of the Uploaded Influence in the context of the Upload Influence Feature, and shall hold harmless Licensor against any third party claim.”
Open citationGenerated from live stance events. Informational only, not legal advice.