Context: Serious business users of Descript Overdub need to have a detailed consent and IP authorisation process when creating voice assets, including clauses for ownership rights remaining with the person whose voice it is (the voice owner), and our obligation to only use their voice for the creation of content that has been reviewed and signed-off by the IP owner. Goal: A fair and legally viable way we have found to use such a tool. Request: Please consider removing the "Voice Identity" feature. I know it's clever but it doesn't work reliably at recognising the person's voice and that costs users money in rework. But even worse than that, it does not interface with the voice owner except at time of its recording -- which means they are unable to retract it at a later date. I ask that you replace it with a means for both parties to upload documents and/voice files in accordance with the agreed process, in the way another user has asked. By doing so you can make mutual trust and robust process the centrepiece, rather than making a meaningless "voice ID" file the sole feature for fair and ethical use. Currently the Voice ID feature doesn't achieve what I presume is the aim, which is to ensure the voice owner understands and consents to creation of a voice asset. We have spent several times more money on rework with voice talent trying to use the "Voice ID" feature than we have spent on the entire Descript product. This is untenable. We have consent documents, which should be attachable to the file -- with the voice owner able to log in and see what files have been made using their voice, and the ability to disable the voice asset.
