Text watermarking will surely cause a degradation in quality writing.

There's is no text watermarking that doesn't cause a decline in writing quality.

Writing quality will surely deteriorate if text watermarking is implemented.

A lot of random-ish numbers go into LLM output and there's a lot of room to hide signals in random-ish numbers.

The real problem is false positives. One false positives is enough to make the whole thing dangerous. The results can't really be acted upon without risking defamation. If you admit that you redistributed someone else's copyrighted work to an AI company that never forgets, it's an admission of distributing copyrighted works.

The law should have at the very least required offline validation tools that cannot track or retain a copy of the documents being checked.