AI Imagery Labeler

ANNOUNCEMENTS AND UPDATES! Thank you for your patience, everyone. Yes, we are back.

We have decided to label this post as "Suspected AI Imagery". Here's how we came to this conclusion, as an example of our process. It takes time and effort, and we try to make our due diligence as much as possible. First, there is no known source for this video. No news coverage. 1/6

At the end of the day what people keep asking for is the ability to know when something is or is not AI so they can make their own informed choices about how they interact with it. Hopefully tools like SynthID become widespread so it doesn't fall on the user to make their own determinations.

Sampling of some of these tools, in no particular order, on the same image. Using these is virtually the same as tossing a coin.

This is why you have to be careful with AI detectors. Some of them are a lot more accurate than others, and none of them are perfect! Our icon was made by two of our artist volunteers. Here is the original concept and some of the iterations they went through.

We've updated so that you can now also make appeals directly through our query page! query.aimod.social Of course users can also still appeal through Bluesky by clicking on the tag in question and then "appeal."

Hey everyone! 👋 I want to share an update on @aimod.social:

A bit more on this. We now believe that: - It depends on which PDS the viewing account is located. - It's transient, as in, it eventually syncs up with time. Additionally: It ALSO happens when we remove a label. It keeps showing for hours or days before eventually disappearing for everyone.

We are seeing a new problem with the labeler. We have no idea if it affects other labelers, but we're wondering if it has anything to do with the temporary suspension that happened on the account. Basically, labels stop showing for SOME people. Example for this post here: bsky.app/profile/did:...

We've been getting some confusion regarding some recent changes happening with the labeler, specifically people being added to lists out of the blue, so here's an update that we hope should clarify things. As always, this labeler is for all of you, so let us know if you run into any issues. 😊

🔎 We're Expanding the AI Mod Team! 🔎 The community has grown so much (thank you! 💙), and we're looking for volunteers to join our moderation team. If you have skills in identifying AI-generated images and a passion for keeping our space transparent, we want your help! Interested? Apply here:

Hey everyone! 👋 We’d love your quick feedback to help guide this account and the work we’re doing. If you’ve got 5 minutes to spare, whether you're a subscriber or not, please fill out this short survey 📝 Your thoughts really matter to us, thank you! 💙

for those who are curious, I did an audit of the total number of accounts for each label on @aimod.social. All of these have been labeled by a human. Our team is amazing. 💙 AI Modified : 1 AI Assisted : 17 Occasional AI Imagery : 1,542 Frequent AI Imagery : 5,876 AI Avatar or Banner : 10,043

Thank you so much to everyone who applied to volunteer with us! We’ve filled the current openings, but we truly appreciate your enthusiasm and may still reach out for future team expansions. Please make sure we can DM you from this account so we can reach you. Thank you again for your support! 💙

Another thing to look out for, besides AI prompters claiming to be human artists, are users who claim things like, "I shouldn't be labeled for AI because I trained my own model." That statement isn't accurate for a couple of reasons. 🧵 1/4

🔎 We're Expanding the AI Mod Team! 🔎 The community has grown so much (thank you! 💙), and we're looking for volunteers to join our moderation team. If you have skills in identifying AI-generated images and a passion for keeping our space transparent, we want your help! Interested? Apply here:

Hi AI Imagery Labeler subscribers! Our report queue continues to grow while our moderation tools, along with Bluesky, are experiencing slowness. Your reports will take longer than usual to be reviewed, but we're doing our best and will gradually catch up! Thank you again for your understanding! 🔎

One of the labelers I subscribe to is @davis.social's @aimod.social, which warns me about users who frequently post AI or AI-modified art, identifies/hides posts with AI imagery, etc. Thanks to Davis & moderator team! New to labelers? See Davis's explanation: publish.obsidian.md/debbieohi/bl...

Hi AI Imagery Labeler subscribers! With the recent influx of users and new subscribers, our report queue has grown significantly. Your reports will take longer than usual to be reviewed, but we're doing our best and will gradually catch up! Thank you for your understanding! 🔎

Hi AI Moderation Service subscribers! If you come across content that you believe to be illegal or NSFW, please send your reports to @moderation.bsky.app Our @aimod.social volunteers are only prepared to review whether content contains AI images, not other categories. Thank you!

Hi AI Moderation Service subscribers! If you come across content that you believe to be illegal or NSFW, please send your reports to @moderation.bsky.app Our @aimod.social volunteers are only prepared to review whether content contains AI images, not other categories. Thank you!

To remove AI “art” from your skyline, sub to @aimod.social 🔎 #noai

There are many reasons to be concerned about genAI. Here are just a few that we've listed in our @aimod.social docs.

Hi AI Moderation Service subscribers! If you come across content that you believe to be illegal and urgent, please be sure to send your reports to @moderation.bsky.app. Our @aimod.social volunteers are prepared only to review whether content contains AI images, not other categories. Thank you!

We want to make sure our labeler is something you can truly rely on, and we know that trust is something we need to earn. If we ever miss the mark, your feedback is essential. Here's how you can appeal if you think we've made a mistake. Don't hesitate to reach out—we're here to make things right!

I do like @aimod.social & appreciate the work of the volunteers who help maintain it! Pls remember most things like labelers & custom feeds aren’t run by the Bluesky devs themselves; they’re made and run by individuals & volunteers trying their best to make this place better for everyone 💖

We wanted to make it easier to understand who we are and how we work as a team, so we created these simple infographics. If you have any questions, feel free to ask!

Is the generative AI bubble showing signs of cracks ? For me an AI-free social network is a feature, not a bug. I like how Bluesky makes it possible to label AI content (example: @aimod.social ) and hope it gets better and stricter on AI-generated content detection. substack.com/home/post/p-...

To date, @aimod.social has labeled 150k posts / accounts for AI imagery.

ANYway I want to give a shoutout to @aimod.social Someone suggested it to me when it first hit Bsky and it really, really does make the difference. It's a labeler service. In Bsky words that means other people help report to this labeler, which labels AI content. You can choose to block outright!

For those curious about how the @aimod.social auto-labeler works and how we make decisions on the team, this is the document we use. When we learn new things and change our methods, the document gets updated as well.

Coming across @aimod.social made me realize how absolutely amazing BlueSky is. Seriously, it's both as easy to use as a Twitter-like centralized service and as customizable as a self-hosted platform à la Mastodon. Anyone is free to create their own moderation filter and share it with everyone!