Meta deletes all AI character profiles on Facebook, Insta after backlash

Meta has close their AI character accounts after the reaction, NBC News and others reported.

While Meta launched these characters in 2023, with AI characters with celebrity avatarsmany online (re)discovered them this week after a recent Financial Times interview with Meta’s VP of Product for Generative AI, Connor Hayes. Hayes mentioned the AI ​​characters on Instagram and Facebook, saying: “We expect these AIs to actually exist, over time, on our platforms, just like the accounts.

“They’ll have bios and profile pictures and they’ll be able to generate and share AI-powered content on the platform … that’s where we see this all going,” he continued.

Meta shut down its celebrity AI avatars last yearbut the non-celebrity AI profiles continued – though many have stopped publishing in 2024, 404 Media reported. U Financial Times The interview, however, made people look for these profiles. What they discovered were offensive depictions of marginalized groups. An example is Meta’s AI character profile “Liv”, who was described as “Proud Black queer momma of 2 & truth-teller”:

Tweet may have been deleted

The customs of life Washington Post columnist Karen Attiah in chats that its creators “admit that they lacked different references”, saying later no Black people were involved in its creation.

In addition to Liv, Meta also created profiles such as “Grandpa Brian“, a black retired businessman, and “Carter”, a dating coach. The discovery of these profiles led to an outroar on X, Blueskyand Meta-properties The threadsNBC News reported, and since publication Meta has removed it all 28 AI profiles announced in September 2023 – both celebrity and non-celebrity.

A Meta spokesperson told NBC News and 404 Media that these profiles were removed due to a “bug” in users’ ability to block profiles:

There is confusion: the recent Financial Times article was about our vision for existing AI characters on our platforms over time, without announcing any new products… The accounts referred to are from a test we launched at Connect in 2023. These were managed by humans. and it was part of an early experiment we did with AI characters. We identified the bug that was affecting people’s ability to block those AIs and delete those accounts to fix the problem.