Usually, these are different things
A hidden Unicode character is a real code point in the text. A zero-width space (U+200B), for example, can be located at a particular position and removed without rewriting the surrounding words. Free AI Wiper's text cleaner reports supported characters by code point, line and column.
A statistical text watermark can be a pattern across word choices made during generation. It may involve no special character at all. Google DeepMind describes SynthID for text as changing token probabilities to create a detectable pattern in generated text. Deleting zero-width spaces would not, by itself, test or remove that kind of signal.
Why the distinction matters
You might find a zero-width space in text written by a person, perhaps because it was copied from another application or used for formatting. Conversely, generated text can contain no unusual Unicode characters. Neither finding is an authorship verdict.
If you need to clean copy-paste artifacts, follow the hidden-character guide and review each character in context. If you need to evaluate a claim that text was AI-written, read what AI detectors can prove. Free AI Wiper does not detect statistical text watermarks or assign an authorship score.