Bot profiles generated by Meta’s AI are not well received

In September 2023, Meta made many of its novel AI chatbots apply celebrity likenesses: Everyone from Kendall Jenner to MrBeast has hired themselves to impersonate AI characters on Instagram and Facebook. The celebrity-based bots were shut down last summer after less than a year, but recently users discovered that several other completely bogus bot profiles are still circulating online – and the response hasn’t been good.

Is “Jane Austen”, “cynical novelist and storyteller”; “Liv” whose bio claims she is a “proud black queer mom of two and a truth teller”; And “Driver” which promises to give users relationship advice. They are all labeled “AI powered by Meta” and the profiles are from when the initial announcement was made. However, the dozen or so AI characters were apparently not very popular: each of them has only a few thousand followers, and their posts receive only a few likes and comments.

“What the hell does an AI know about dating????” reads one of the latest comments on the AI ​​dating trainer bot profile. “This isn’t just virtual blackface, this is just weird,” a commenter wrote in a post on Liv’s page.

Another point of anger is that there doesn’t seem to be a way to block bots via typical channels: there’s no option to block or restrict profiles.

Many AI bots have not shared novel content on their networks since early 2024, and it is unclear how (or if) users found and engaged with these profiles over the past year. Last week Financial Times. reported that Meta predicts a future where social media platforms will be filled with AI bots.

“We expect these AIs to exist on our platforms over time, much like accounts do,” Connor Hayes, vice president of product for generative AI at Meta, told the outlet. “They will have biographies and profile photos and will be able to generate and share AI-powered content on the platform. . . that’s where we see it all happening.”

However, Meta confirmed that the bot profiles have existed since 2023 and are part of an early “human-managed” experiment.

“Recently Financial Times. the article was about our vision for AI characters existing on our platforms over time and did not announce any novel product,” said Meta spokeswoman Liz Sweeney Edge in e-mail. Sweeney said the company has identified a bug affecting users’ ability to block accounts and that profiles are being removed to fix the issue.

The idea of ​​deliberately flooding social media with bots is absurd at first glance, but it’s consistent with the way Meta promotes generative AI tools. Anyone in the US can create their own chatbot, and the concept is that users will be able to send their bots to their place to talk to their followers. Chatbot services like Character.ai have gained popularity in the past year among people looking for a digital friend or just a way to pass the time, but artificial intelligence companies are also facing lawsuits accusing them of posing a risk to users. including children.

Update, January 3: Added a response from Meta spokeswoman Liz Sweeney.