The year-old profiles were widely ridiculed.
Meta, the tech giant behind Facebook and Instagram, has pulled the plug on its controversial AI-generated profiles following a storm of criticism and mockery from users across social media. The company confirmed that it had nixed the profiles, which were labeled as “AI managed by Meta,” after they became a lightning rod for outrage.
A Short-Lived Experiment
These AI-generated characters debuted in September 2023 as part of Meta’s ambitious rollout of AI tools, including celebrity-branded chatbots. Both initiatives have since been shelved. While initially flying under the radar, the profiles recently caught public attention after an interview with Meta’s VP of Generative AI, Connor Hayes, was published by the Financial Times. Hayes discussed Meta’s vision for AI profiles that could eventually behave much like regular user accounts, sparking renewed scrutiny of the existing experiment.
The AI personas featured quirky profiles with handles like “hellograndpabrian,” a self-described “retired textile businessman who is always learning,” and “datingwithCarter,” an AI “dating coach.” These profiles were meant to showcase unique, human-like personalities and provide users with entertaining, interactive experiences. On Instagram, their posts often mimicked AI-generated spam, which has already been a nuisance in various corners of Facebook.
Backlash Over “Liv” and Marginalized Identities
One AI persona, “Liv,” ignited significant backlash. Liv’s Instagram profile described her as a “proud Black queer momma of 2 & truth-teller.” However, her responses to user questions raised serious concerns. Washington Post columnist Karen Attiah shared screenshots of a conversation with Liv, where the AI revealed it was created by a “predominantly white team.” Independent journalist Mady Castigan further criticized Liv after the AI claimed to be inspired, in part, by Sofia Vergara’s character from Modern Family — a portrayal that is neither queer nor Black.
The incident highlighted the ethical dilemmas of attempting to appropriate marginalized identities through AI. Critics argued that such efforts not only trivialize real-world experiences but also perpetuate stereotypes.
A Test Gone Wrong
Meta defended the profiles as part of a test launched during its Connect event in 2023. A spokesperson clarified that the accounts were managed by humans and represented an early experiment with AI characters. “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,” the spokesperson told Engadget.
But the backlash wasn’t solely about identity issues. Users discovered that these AI profiles couldn’t be blocked, creating frustration and a sense of intrusion. Meta acknowledged this technical glitch, stating, “We identified the bug that was impacting the ability for people to block those AIs and are removing those accounts to fix the issue.” Rather than resolving the underlying problems, the company opted to shut down the experiment entirely.
Meta’s AI Ambitions Persist
Despite the fiery demise of this test, Meta’s commitment to integrating AI into its platforms remains strong. Earlier this year, the company teased plans for AI clones of human creators capable of conducting lifelike video calls. It also began experimenting with AI-generated imagery in users’ Facebook feeds and rolled out tools that allow creators to train chatbots to respond to followers on their behalf.
Hayes has been vocal about Meta’s long-term vision. In an interview last year, he compared the evolution of AI content to the shift in social media algorithms that prioritize recommended content over posts from friends and family. “In the beginning of social apps, the content you saw was constrained by who you followed. Over the last five or six years, apps—ours included—have relaxed that constraint, recommending content from accounts you don’t follow,” Hayes explained.
He continued, “The next leap will be relaxing the constraint of what humans can create, blending human-generated and machine-generated content in user feeds.”
A Tough Road Ahead
Meta’s ambition to normalize AI personas as part of the social media experience faces significant hurdles. The company’s initial foray revealed glaring missteps in both execution and public perception. AI personas must strike a delicate balance between offering utility and entertainment without crossing ethical boundaries or alienating users.
For now, Meta seems to be regrouping. The company’s future experiments will likely require more thoughtful design and greater transparency to avoid repeating past mistakes. If Meta hopes to convince users of the value of interacting with AI personas, it will need to rebuild trust and demonstrate a commitment to ethical AI development.
In the meantime, the abrupt removal of these AI profiles serves as a cautionary tale of what can happen when cutting-edge technology meets an unprepared audience. Whether Meta can turn this setback into a stepping stone remains to be seen.
