By Ben Abrams x The Big Magazine
The company’s creation of AI-generated Instagram accounts sparked a viral controversy, resulting in a blackout of search results.
Meta is removing the Facebook and Instagram profiles of AI characters it developed over a year ago after users rediscovered these profiles and interacted with them, leading to viral screenshots of the conversations.
Prior to Friday, those AI accounts did not garner much notice. Once they did, controversy ensued, prompting Meta to remove the accounts and limit search results for their usernames.
The company initially launched these AI-powered profiles in September 2023, but discontinued most of them by summer 2024. Meta, the parent company, discreetly maintained several artificial intelligence accounts on Instagram and Facebook. These accounts were initially introduced alongside a collection of official celebrity AI characters. The AI-powered personalities shared some AI-generated images and could be engaged in conversation through direct messages.
Nonetheless, a few characters persisted and attracted renewed interest after Meta executive Connor Hayes mentioned to the Financial Times late last week that the company intended to introduce more AI character profiles.
The automated accounts posted AI-generated pictures to Instagram and answered messages from human users on Messenger.
Following the viral spread of the profiles, they started vanishing. Users observed that these profiles couldn't be blocked, which was described as a bug by Meta spokesperson Liz Sweeney. She explained that the accounts were human-operated and were part of a 2023 AI experiment. The company took down the profiles to address the bug that stopped users from blocking the accounts, according to Sweeney.
In the wake of the Financial Times article, users resurfaced some of the 2023 AI characters, particularly one called “Liv” depicting a “Proud Black queer momma” who solicited messages from human users.
In July 2024, Meta discontinued its AI characters featuring celebrities and introduced AI Studio, a tool that allows individuals to create personal AI characters, which can also be utilized by others via the messaging features on Meta’s social media platforms.
When Washington Post columnist Karen Attiah began conversing with “Liv,” she shared a series of screenshots of its responses, which included the AI account stating, “My creators admitted they lacked diverse references.”
It mentioned the supposed racial and gender composition of the development team behind the chatbot, noting it didn’t include any Black people, and what the account claimed to be the name of its developer. Meta has not confirmed the authenticity of the AI character’s claims. It remains uncertain if the name provided by the AI account belongs to a real Meta employee or is a fictional character.
“There seems to be some misunderstanding: the recent article in the Financial Times discussed our long-term vision for AI characters on our platforms, not the launch of a new product,” Sweeney clarified in a statement. “The accounts mentioned were part of a test we initiated at Connect in 2023. These accounts were controlled by humans and were part of an initial experiment involving AI characters. We found the bug that affected users' ability to block these AIs and are removing those accounts to resolve the problem.”
While these Meta-generated accounts are being removed, users still have the ability to generate their own AI chatbots.
Follow us on Instagram @TheBigMagazine
Comments