Spam users with AI generated content

Curators
1

I came across a user where probably all of their observations use AI generated images. I flagged them as a spammer (they have 70+ violations this month), which caused their account to be suspended. However, their 66 existing, dubious observations are still visible on their profile. I assume that observations from suspended users no longer show up in search results, and are only visible on their profile to curators?

I noticed that the iNaturalist Community Guidelines do not list spamming as a suspendable offense. This seems like an oversight, and clearly does accurately reflect how the Admin Tools work. Also, I was not able to find reference to the Admin Tools in the Curator Guide.

Out of curiosity, has anyone else encountered a user with entirely AI generated observations? This was my first time, but I imagine it will only become more common.

As an aside, the user made the mistake of uploading an AI generated robber fly, and because I had recently tested Nano Banana 2’s ability to generate Asilidae images, I recognized that garbage immediately. I recommend a similar approach for other curators so they know what to look for in their taxa of interest.

2

I am aware that AI can now occasionally create convincing AI Chironomidae images to the untrained eye. Which is most people. AI generated content like that is a violation, so suspension is called for.

3

There is a relatively new flagging reason specifically for AI-generated content: https://www.inaturalist.org/blog/118284. This seems more appropriate to me than flagging it as spam. That being said, if the system flags you as a spammer, you will lose your account anyway.

I took a look at the flag overview to see which user you were referring to and whether I (as a non-curator) could still see their content. The answer is: No, I can’t. If I click on their username, I get a notice that the user has been suspended. If I click on the photo that was flagged as AI-generated content, I get a notice that the image was flagged, and an ‘AI content’ placeholder image appears instead of the actual photo.

I haven’t encountered a user who exclusively uploaded AI images yet. However, I would generally disagree with your approach regarding Nano Banana 2. In this thread here: https://forum.inaturalist.org/t/gemini-3-5-flash-species-accuracy-can-we-still-spot-the-ai-image/79264, I generated AI photos using the latest version of Google Gemini Pro (and thus Nano Banana 2) in the top post (https://forum.inaturalist.org/t/gemini-3-5-flash-species-accuracy-can-we-still-spot-the-ai-image/79264?u=fgallhuber) and further down (https://forum.inaturalist.org/t/gemini-3-5-flash-species-accuracy-can-we-still-spot-the-ai-image/79264/39?u=fgallhuber). You really can’t tell the difference anymore—especially with the examples further down. If you know how to properly prompt the tool, you can get highly detailed and accurate images ;)

4

I doubt the AI will be sending out accurate genitalia images anytime soon. There are 1000s of chiro with no images on the internet.

5

Thanks for looking out, but please only use the spam flag for actual spam.

The account should be flagged as spam automatically once three pieces of its content are flagged as spam.