The electronic frontier foundation explains why filters are incompatible with Fair Use:
https://www.eff.org/deeplinks/2020/0...itute-fair-use
Quote:
Content ID works by having copyright holders upload their content into a database maintained by YouTube. New uploads are compared to what’s in the database and when the algorithm detects a match, copyright holders are informed. They can decide whether to monetize someone else’s video for themselves, mute the audio, or take it down. Users whose videos are hit with Content ID can dispute the match—chancing the claim being converted to a strike—or alter their video in some way that releases the claim.
Content ID makes matches based on seconds of matching audio or video. In other words, it doesn’t just make matches when a whole thing has been copied and uploaded. It makes matches when just a short clip is found. And short clips are often present in videos making fair use.
|
Quote:
Videos critiquing a film or song are going to include clips from that video or song. It makes the point stronger. In the same way that high school English classes teach students to put quotes in their essays to make their point stronger, people working in visual and audio formats do the same thing.
Moreover, fair use gives people the legal right to use copyrighted material for purposes like commentary and criticism without having to get permission or pay the copyright holder. And fair use isn’t bound by a specific number of seconds. It’s bound by whether what was used was needed for the point being made.
|
Thefe are legal standards for what is and isn't Fair Use but the standards are based on context and rationality; things whether the content use is transformative, whether it can substitute for the full product, the intent behind the use, etc.
Some cases are easy to indentify but many will challenge even educated, intelligent lawyers and judges. Few if any are reducible to easily coded algorithms.
With today's tech, you can have Fair Use or you can have automated filters.
Not both.