Meta Implements New AI Disclosure Requirements

Meta’s taking more steps to ensure that its users are aware of AI-generated content in its apps, with a new AI-generated label option now showing up within the post composer flow, as part of its expanded AI disclosure requirements.

As you can see in this example, posted by app researcher Muhammad Jalal (and shared by Matt Navarra), Meta’s now rolling out a new tag that it will require users to activate when they upload AI-generated content.

This is in addition to Meta’s own AI detection tools, which will append its own “Made with AI” labels to content where Meta detects AI image indicators.

The manual post tags will also result in the same “Made with AI” disclosure being displayed on your post, which will ideally ensure that Facebook and IG users are aware that the image is not real, and help to alleviate confusion about what people are seeing.

Which is a good thing, because right now, AI-generated images are causing significant problems for Facebook users.

As reported by 404 Media this week, an increasing number of Facebook pages, in particular, are now posting increasingly disturbing AI generated images in order to fish for likes.

Those images range from ridiculous junk creations in seemingly remote villages:

To drowning kids:

Any level of scrutiny will reveal an array of errors with these AI-generated pictures, but as you can see, that hasn’t stopped them from garnering hundreds of thousands of likes and comments from seemingly unaware Facebook users.

Which is why proper disclosures are needed. And while the new AI tags will provide a level of transparency, they are small, and given that so many users didn’t detect that these images were fake, it’s likely that many will also miss these notes as well.

Which could become a major problem for Meta, as more scammers and spammers seek to use AI images as a driver of engagement.

Generally, these scammers seek to boost their Pages with tactics like this, in order to then onsell the Page to others, with an established large audience, or they look to post spam links or propaganda, as well as their usual posts, helping to get more reach for those messages.

As such, I would expect that Meta will be looking to strongly enforce its new AI disclosure rules, which came into effect this week, while I would also anticipate further reach penalties for such, when it’s either disclosed or detected.

And clearly, Meta needs to do something, otherwise your Facebook feed is going to be all AI-generated junk within a few months.