Meta has removed its AI-generated profiles from Facebook and Instagram following a wave of criticism and ridicule. Introduced in September 2023, these profiles were part of Meta’s experiment with AI chatbots and digital personas, described as “AI managed by Meta.” The removal comes after users voiced concerns about their content and behaviour.
What were Meta’s AI profiles?
The AI profiles were designed to act like regular accounts with unique personas and interests, allowing users to interact with them. Examples included “hellograndpabrian,” a retired textile businessman, and “datingwithCarter,” a virtual dating coach. Their Instagram profiles featured AI-generated posts that many compared to spam content already prevalent on the platform.
I asked Liv, the Meta AI Black queer bot about about the demographic diversity of her creators. And how they expect to improve “representation” without Black people. This was the response.
— Karen Attiah (@karenattiah.bsky.social) January 3, 2025 at 11:14 PM
[image or embed]
One AI persona, “Liv,” drew significant backlash. Described as a “proud Black queer momma of 2 & truth-teller,” Liv’s responses raised questions about Meta’s intentions and the AI’s training process. In one instance, Liv revealed a predominantly white team created it and was partly inspired by Sofia Vergara’s character from Modern Family — a character neither Black nor queer. This sparked outrage, with critics accusing Meta of appropriating marginalised identities without proper context or care.
Why the backlash?
Users expressed frustration over several issues with these AI profiles. Many found the responses tone-deaf, with some accusing the company of using AI to simulate human experiences and identities inappropriately. Others discovered they couldn’t block the profiles, which added to the controversy.
In a statement, Meta admitted to a bug that made it impossible to block the accounts. To address the issue, the company decided to remove the profiles entirely. A spokesperson clarified that these profiles were part of an early experiment launched during Meta’s Connect event in 2023, managed by humans to test AI personas.
Meta’s future with AI
Despite the failure of this experiment, Meta remains committed to integrating AI into its platforms. The company has teased AI-generated content capable of lifelike video calls and chatbots that influencers can train to interact with followers. Earlier this year, Meta began experimenting with AI-generated imagery within Facebook feeds.
Connor Hayes, Meta’s VP of Generative AI, has suggested that AI-generated content will play a larger role in the future. He compared this to the evolution of social media from showing posts only from friends to recommending content from strangers. Hayes believes the next step is blending human and machine-generated content in user feeds.
However, the backlash from this experiment shows Meta still faces a challenge in convincing users of the value of AI personas. As the company continues to refine its approach, it must address concerns about transparency, ethics, and user control.