SlopOrNotAI (@SlopOrNotAI) header
SlopOrNotAI (@SlopOrNotAI) profile picture

World's first AI content detector for text and images that runs on-device on iPhone - fast, accurate, and fully private. https://nitter.cf/t.co/kbveeinGm1

Europe
Joined June 2025
Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)Photo by SlopOrNotAI (@SlopOrNotAI)
SlopOrNotAI retweeted
It doesn’t work. Removing C2PA is trivial. You don’t need a skill for that. Command-line tools already exist. The real test is SynthID and other steganographic watermarks embedded in the image. Generate an image with ChatGPT, remove the visible watermark, run it through @SlopOrNotAI, and see what happens.
8
9
1
25
5,926
SlopOrNotAI retweeted
The hidden watermark used by @OpenAI’s GPT-Image-2 has been extracted. It appears to be this dense pattern, normally invisible to the human eye. This kind of watermark is typically used to identify AI-generated images later, even if there’s no visible label attached. Could it be related to the model’s infamous noise artifact issues? We don’t know yet, but I doubt it.
I extracted the gpt-image-2 watermark! testing it on different types of images now
58
36
13
750
166,409
The Nano Banana 2 leaks are insane. You can’t even tell what’s real anymore. But I found this AI content detector that instantly flags Nano Banana 2 images. (Meet Slop or Not👇)
Nano banana 2 seems to be far superior to its predecessor. What is particularly remarkable, however, is the speed with which the second model has improved. Just think of Genie 3 – will we also see Genie 4 very soon?
16
7
2
144
35,983
SlopOrNotAI retweeted
🚨 New launch: @SlopOrNotAI hits the App Store today - the world's first on-device AI content detector for iPhone! Catch AI-generated text and images right away, with all data staying private on your phone. In a time of rising AI fraud, scams, and misinformation, dodge catfishes on dating apps by checking profiles, spot fake news on social media, verify product reviews before buying, or confirm essay originality for school. Unlike cloud-based detectors that leak data, ours uses custom models locally for both. No logins, no telemetry, no logs. Free download with daily checks - It’s time to fight back against AI Slop!
1
2
8
377
SlopOrNotAI retweeted
The latest beta of @SlopOrNotAI now supports detecting images generated by Nano Banana.
1
2
5
386
SlopOrNotAI retweeted
Not to mention, ChatGPT generated ghibli slop.
1
1
7
723
SlopOrNotAI retweeted
And of course, text generated by ChatGPT and other thousands of other text AI apps.
1
1
5
324
SlopOrNotAI retweeted
Also works on reliably on content from deepfake generation sites.
1
2
5
1,052
SlopOrNotAI retweeted
If you’re worried about the oncoming onslaught of AI generated images, we’ve been working on an on-device AI app that detects them reliably. BTW, also works for text slop and slop images from other models and providers. #WWDC2025
4
4
25
2,814