Meta has deleted the Facebook and Instagram profiles of AI characters it created more than a year ago after users rediscovered parts of the profiles and joined conversations, and screenshots of them went viral. are.
The company first introduced these AI-powered profiles in September 2023, but retired most of them by summer 2024. However, after meta executive Connor Hayes spoke, a small number of characters remained and received renewed interest. financial times It was announced late last week that the company plans to roll out more AI character profiles.
“We expect these AIs to actually live on our platform over time, just like accounts,” Hayes told the FT. The automated account posted AI-generated photos on Instagram and responded to messages from human users in Messenger.
These AI profiles include Liv, who describes herself as “a proud black queer mom of two and a truth teller,” and someone with the account handle “datingwithcarter” who describes herself as a relationship coach. That included Carter explaining. His profile says, “Message me for a better date.” Both profiles include a label indicating that they were managed by Meta. The company released 28 personas in 2023. Everything was closed on Friday.
Conversations with the characters quickly went off topic as some users asked questions such as who created and developed the AI. For example, Liv said: creator team Blacks were not included, and the majority were white and male. This was “a pretty obvious omission given my identity,” Bott wrote in response to a question from Washington Post columnist Karen Attia.
Within hours of the profile going viral, it started disappearing. Users also noted the inability to block these profiles, which Meta spokeswoman Liz Sweeney said was a bug. Sweeney said the accounts were controlled by humans and were part of an AI experiment in 2023. Sweeney said the company removed profiles to fix a bug that prevented users from blocking their accounts.
After newsletter promotion
“There is some confusion. The recent Financial Times article is about our vision for the long-term presence of AI characters on our platform, not announcing a new product,” Sweeney said. said in a statement. “The accounts referenced are from tests we started with Connect in 2023. They were managed by humans and were part of our early experiments using AI characters. , we identified a bug that was impacting people's ability to block these AIs and are removing those accounts to resolve the issue.”
These meta-generation accounts have been removed, but users can still generate their own AI chatbots. User-generated chatbots promoted to the Guardian newspaper in November included a “therapist” bot.
Once you started a conversation with a “therapist,” the bot suggested several questions to ask to get started, such as “What can I expect from the session?” and “What is your approach to treatment?”
“Through gentle guidance and support, I help my clients increase self-awareness, identify patterns and strengths, and cultivate coping strategies to overcome life's challenges,” 96 followers and 1 post A bot created by an account with answered.
All of Meta's chatbots include a disclaimer that some messages may be “inaccurate or inappropriate.” However, it was not immediately clear whether the company moderated these messages or ensured that they did not violate its policies. When a user creates a chatbot, Meta will guide them on the type of chatbot they should develop, such as “loyal confidant,” “enthusiastic listener,” “tutor,” “relationship coach,” “sounding board,” or “sounding board.” I would like to suggest some. The all-seeing astrologer.” A loyal best friend is described as “a humble, loyal best friend who is always there to support you behind the scenes.” Relationship coaching chatbots help bridge the “gap between the individual and the community.” Users can also create their own chatbots by: Character description.
Courts have not yet answered what liability chatbot creators bear for what their artificial companions say. US law protects social network creators from liability for what their users post. But a lawsuit filed in October against Character.ai, a startup that develops a customizable role-playing chatbot used by 20 million people, alleges that the company claims it has an addictive product that encourages suicide among teenagers. is claimed to have been designed.
