I think a better approach to outright blocking the infringing, or presume to be infringing content, would be developing software that identifies the infringing

I think a better approach to outright blocking the infringing, or presume to be infringing content, would be developing software that identifies the infringing audio and removes it using sound cancellation. This would require the claimant to submit a sample of the infringed content and an algorithm to add a layer of opposite polarity audio or alteration of the actual original layer. The same would be more difficult with video, but could be done eventually.: This approach would require a great amount of initial investment. It would have to be designed to deal with frequency shifted audio as well.