Sure, it's first tool that seems to actually shows specific left-over code so congrats, I guess, you can find small % of fics done specifically with this one genAI.
Whatever benefit you imagine coming from it, the negative repercussion are gonna be astronomically worse.
Cause what this thing doesn't tell you it WHY and HOW the Claude was used.
You have no clue! You don't know! The makers don't know! But you can bet that people online will not do any kinda of research - the person who prompted the entire fic and ended with 1k+ Claude code instances found in 20k fic will be harassed and bullied just as badly as someone whose English is second language and wrote something in their native one and used Claude to translate, because they thought if every translator and spell-checker is adding genAI then there is no difference, so they ended with 2 instances among 200k word fic, and just as badly as 12 year old who just wrote their first fic and wanted it to sound cooler and this fun website was advertised everywhere and they had no clue what genAI is or why it's unethical to use, because copyright infringement or environmental impact even are in their vocabulary yet, and just as badly as someone with dyslexia/hand tremors/ADHD who put it there to ask it to fix any typos because their uni tells them to use it, so it can't be bad.
[Also, what does '1 instance mean'? Does it mean one copy paste? Does it mean one paragraph? Is this specific 50 instances a 50 cherry-picked sentences, or is it 50 multiple-page-in-Word-worth long fragment? We don't know, the skin makers didn't explain, just posted the scary big numbers.]
There will be no nuance to be had among the guilty - and don't even get me started on the amount of just accusations that will happen. And there is no way to prove that your chapter wasn't edited - AO3 does not show it in any way, shape or form.
The amount of people I've seen who complain about the skin not working right and tagging wrong works (someone hilarious got it tagging every single work on AO3) because they're tech illiterate, or people already claiming 'xyz fic sound genAI so I know it definitely, they must've just removed the trace so it doesn't detect it' or outright telling people glad the works they checked came clean 'don't trust it, that means nothing, they could've just cleaned it'.
Literally witch trial logic - if it shows Claude code you're guilty cause it's there, if it doesn't you're still guilty and must've just cleaned it up. You cannot win.
You think it doesn't happen already? That people aren't blocked and judged based on ~vibes~ alone - that this method of judging how something 'sounds' - something that is known sound very similar to people on the spectrum because of the neuro-divergence AND more recently to black people it was trained and fixed by- isn't approved of?
Also, humans are shit at detecting genAI. Not, you're not better, writing is not the same as art, listen to literal wikipedia, I am begging you.
This is TERRIFYING. It no longer matter this tool is unprecedentedly reliable or the small benefit it might bring: it became tool of harassment within HOURS of being released. Can you imagine how much worse it's gonna become?
You hate someone? Accuse them of using genAI! It takes 5 second to fake a screenshot, you can edit source code in notepad and nobody's gonna check for themselves before joining the mob online - they never do.
By the time an author proves themselves innocent, they'll have been flooded with hate and harassed at best or doxxed or attacked in real life at worst.
However theoretically useful this tool might me: it was not released to help anyone, or every in good faith. It was released to make send a hate mob and stoke the flames of harassment.
If they cared for transparency, they would've gone to AO3 to let them use it as tool to improve implementing genAI content tagging.
That would've proposed to AO3 that they request people to tag genAI content and offered this as a reliable tool of checking compliance, without outing the tools' function and teaching people how to hide better.
Instead they made a skin to create a hate campaign for a very queer fandom, blasting two users by username in their main tweet, even when they knew it would result in genAI users learning how to hide their tracks immediately. They know this for certain, because people immediately wend to call for witch-hunts finding as many authors as possible with their invention before they wisen up to it. The skin makers made no push-back against it.
Also, I will never take it in good faith given they only called out fics by name for Heated Rivarly fandom, cause that's the fandom they personally like. Apparently they care 'so much' about transparency only when it's personal for them, otherwise they give a perfunctory 1 fic title in random 4 big fandoms to look less like haters and call it a day.
Just to illustrate my point: this is from 3 months ago. It's already >this bad<. This tool will fic nothing and only make things worse.
Also, the 'do not harass pls uWu' is not even sign of good will. Forthose unaware, it's absolutely a legal disclaimer.