Meta has nuked a bunch of its AI-generated profiles from Facebook Instagram, the company confirmed, after the AI characters prompted widespread outrage and ridicule from users on social media.
The AI-generated profiles, which were labeled as “AI managed by Meta,” launched in September of 2023, rolling out alongside the company’s celebrity-branded AI chatbots (also discontinued). Meta doesn’t seem to have updated any of these profiles for several months, and the pages seem to have been largely unnoticed until this week, following an interview published by the Financial Times with Meta’s VP of Generative AI, Connor Hayes.
In the interview, Hayes spoke about the company’s goal to eventually fill its services with AI-generated profiles that can interact with people and function “kind of in the same way that accounts do.” Those comments brought attention to the extant fMeta-created AI profiles and, well, users were not exactly impressed with what they found.
With handles like “hellograndpabrian,” a supposed “retired textile businessman who is always learning” and “datingwithCarter,” an AI “dating coach,” the chatbots were meant to showcase “unique interests and personalities” for users to chat with. On Instagram, their profiles also featured AI-generated posts that, as 404 Media noted, looked a lot like the AI spam that’s become prevalent in many corners of Facebook.
content posted by ” src=”https://s.yimg.com/ny/api/res/1.2/Ipbn3xEgkhkHO4u4PE2sSQ–/YXBwaWQ9aGlnaGxhbmRlcjt3PTk2MDtoPTU3NQ–/https://s.yimg.com/os/creatr-uploaded-images/2025-01/f15b8890-ca10-11ef-afbf-3f62277a043d” height=”575″ width=”960″/>
Meta
An AI persona called “Liv” sparked particular outrage. The Instagram profile identified “Liv” as a “proud Black queer momma of 2 & truth-teller.” Washington Post columnist Karen Attiah posted a series of screenshots in which she interrogated “Liv” about how Meta trained the AI, with “Liv” sharing that it was created by a “predominantly white team.” Independent journalist Mady Castigan posted another conversation in which “Liv” said that its creators had been inspired in part by Sophia Vergara’s character from Modern Family, a character that is neither queer nor Black.
“There is confusion: the recent Financial Times article was about our vision for AI characters existing on our platforms over time, not announcing any new product,” a spokesperson told Engadget. “The accounts referenced are from a test we launched at Connect in 2023. These were managed by humans and were part of an early experiment we did with AI characters.”
Beyond sparking ridicule for their responses and attempts to appropriate marginalized identities, users found the AI profiles were impossible to block, for reasons unknown. Rather than fix the issue, Meta’s solution was to kill the experiment entirely. “We identified the bug that was impacting the ability for people to block those AIs,” a spokesperson said, “and are removing those accounts to fix the issue.”
While this trial run has gone up in flames, the company doesn’t seem to be abandoning its plans to bring more AI-generated “characters” to its apps. Earlier this year, the company teased AI clones of human creators capable of holding lifelike video calls. Creators can already train their own chatbots to respond to followers on their behalf. Meta also began experimenting with inserting its own AI-generated imagery into users’ Facebook feeds.
In an interview last year, Hayes told me that Meta likely will become more “proactive” about surfacing AI-generated content over time, comparing it to the shift from showing recommended content instead of posts from people you follow.
“In the beginning of social apps … the corpus of stuff that you could see on a given day was sort of constrained by who you followed or were friends with. And over the last like, five or six years, a lot of apps — ourselves included — have moved to, you know, relax that constraint and start recommending content from accounts you don’t follow.
“I think probably the next leap that’s going to happen there is relaxing the constraint of what humans can create, and actually getting to feeds of content that are a combination of things that, you know, humans have created, but also that are entirely machine generated.”
It may still be awhile before Meta fully realizes that vision. But if the reaction to its early experimentations is any indication, the company still has a lot of work to do to convince people AI personas are worth interacting with in the first place.
