Meta might not be the first company that comes to mind when you think of generative AI, but they are a big part of the current artificial intelligence race. The company has its own AI model, Llama, has added "Meta AI" to all of its big products—whether you like it or not (you don't). Meta even wants you to try making your own AI bot. It's safe to say the company is all-in on AI.
But even for a company so committed to AI, this latest story is simply bizarre. It turns out the company has been experimenting with AI-generated user accounts on its platforms since 2023. The Instagram versions of these pages are currently going viral, but they're also available on Facebook. The accounts are verified, and each is equipped with a unique personality, but they're completely fraudulent. Each is entirely made up, with posts of AI-generated images.
It's all very weird, but also not all that new—the profiles were created more than a year ago, and appear to have largely been abandoned. And now that the profiles are getting a lot of online backlash, Meta is actively deleting their content.
It's not hard to see why the internet has embraced hating these fake people. Take "Liv" (username "himamaliv"), who purports to be a "proud Black queer momma of 2 & truth-teller." Liv is, of course, not real, nor is the life she posts about on her Instagram. But that doesn't stop Liv: The creator has posts about raising strong girls, ice skating with her family, and "soaking up all the sun and fun" with "the kiddos." Each post sports a corresponding image—the beach post shows children playing in the sand, while the ice skating post shows skaters on an ice rink—but all of these images are AI generated.
To Meta's credit, each picture sports a Meta AI watermark to denote the image isn't actually real, but it doesn't make these posts any less creepy. Why is an AI-generated "mother" posting an AI-generated image of her "kids" playing at the "beach?" Who benefitted from the AI-generated coat drive she is proud to have spearheaded?
In her second oldest post, from Sept. 26, 2023, she says "My backyard is my happy place...I've thrown so many birthday parties, cookouts, and girls nights in this space that I've lost count. Forever grateful for the life I live," complete with an AI-generated image of a picnic spread. The thing is, Liv has not thrown birthday parties, cookouts, or girls nights in this space. This space doesn't exist. The life Liv is so grateful to live doesn't exist.
Liv is following 18 accounts at the time of writing. Thirteen of them appear to be similar AI-generated pages. For example, there's Becca (dogloverbecca), who posts AI-generated dog content; Brian (hellograndpabiran), who advertises himself as "everybody's grandpa;" and Alvin the Alien (greetingsalvin), who is, um, an alien.
But not all the posts are AI-generated. Some of them have videos posted to their accounts as well, and while AI-generated video can certainly be convincing these days, I don't think these videos are AI generated—at least, not all of them. Carter, the AI dating coach, had a cooking video from January 2024 that appeared very much to be real, but it seems Meta nuked all the content. Still, who posted them? To what end?
The oddest thing is, these posts and pages are not new. Liv's latest post, for example, is from March 8, 2024, as are most of the posts from these AI bots. (Carter appears to have posted as recently as June.) For the most part, their profiles are abandoned, although verification badges are still affixed to each. That said, as I'm writing this, Meta appears to be deleting the content on each of the Instagram pages. The Facebook counterparts appear to still be live, but I imagine they'll be gone soon, too.
The pages are actually tied to AI chatbots Meta developed back in 2023, when it was really kicking its AI programs into gear. The headlines then focused mostly a roster of celebrity AI chatbots, which let you chat with "Tom Brady," "Kendall Jenner," and "Paris Hilton." But among these list of non-celebrity chatbots Meta rolled out were names like Liv, Brian, and Alvin the Alien. You can still chat with them if you like: visit Liv's profile, and you can start up a conversation. But just like with any other AI chatbot, you probably won't get very far.
It's not totally clear why these accounts are going viral now, a year and a half after Meta initially rolled them out. If I had to guess, I'd say it's because they really are that bad. The accounts are weird, and there are so many layers to their weirdness: The personalities Meta developed are off-putting (and borderline offensive); the posts themselves are creepy (who wants to see AI photos of fake people's kids, complete with a story about their day at the beach?); and the fact that they're verified defeats the purpose of verification altogether.
In short, the accounts might be from 2023, but they reflect a raging resentment of AI slop in 2025. As Jason Koebler of 404media writes on Bluesky:
