>> One possible best practice would be to allow authors to specifically opt out of use of their output for ML training. That is legally the default. Creators o

>> One possible best practice would be to allow authors to specifically opt out of use of their output for ML training. That is legally the default. Creators own their copyrights. In many cases it is made explicit with a creative commons non-commercial use license. Remember, without a license you get nothing commercial - except the nebulous fair use. The real problem here is companies thinking they can consume large amounts of material and works simply because they can see them on the internet and obfuscate them by combining together.