I'd like to spread awareness, but I'd also like to warn that not every comment that looks this way was bot-generated. I've read comments like these from real users and it would be a huge mistake to report someone based on little evidence. It's good to be aware, but please be careful of your actions too. It might just be a legitimate user.
Moving on, I would like to know more about the why of the story. The explanation is plausible, but I don't know... I'm well-versed in IT so I know how bots and AI work. Generating traffic to appear legitimate is a thing (without traffic the "story writer" would be more suspicious, as there's no other activity such as reading, liking and commenting), but only AO3 admins could see it, and the comments don't lead to anything harmful.
The more pressing issue - does that mean that our stories are being fed to the AI to learn? Because if not, there's really nothing to worry about; comments aren't hurting anyone. But stories -- something someone put hours and hours into creating -- I would be worried about that.
Please feel free to spread awareness and add new information to the topic! Hopefully
PSA: bot comments are taking over ao3
The above examples have been provided with the authors' permission to demonstrate what these look like.
Basic rundown:
They are all 3 sentences long
Perfect grammar, capitalization, and punctuation
Like absolutely flawless English teacher-style writing with only a single exclamation mark, ever
No mentions whatsoever of character names, settings, situations, or anything that could be tied to the story
The usernames may be identical to people who exist on ao3, but the name is not clickable, and no profile is associated with it EXCEPT when you directly search for that name. What this means: the comments come from an unregistered (not logged in) reader, bots scrape the site for real usernames, attach that to the comment, and post
Please spread the word about this so authors can filter comments and report them accordingly
There has been some speculation about why this is happening at all, and the best guess is that this is a feature that AI-training story-scraping tools are implementing to try and make their browsing traffic look legitimate