Illustration by Tag Hartman-Simkins/Futurism. Source: Mike Winkelmann/Chanda Khanna/AFP via Getty Images
Mark Zuckerberg has been accused of doing a number of shoddy things on behalf of Meta’s bottom line, but the latest allegations could ruin the rest.
According to internal Meta emails and messages filed in a New Mexico state court case that were made public this week, Zuckerberg himself signed off on allowing minors to access Meta’s AI chatbot companions, even though the company’s security researchers warned that they Posed the risk of engaging in sexual conversations.
Per reutersThe lawsuit alleges that Meta “failed to stop the flood of harmful sexual content and sexual advances directed at children” on Facebook and Instagram. “Zuckerberg-run Meta rejected the recommendations of its integrity staff and refused to put in place appropriate guardrails to prevent children from having sexually exploitative interactions with its AI chatbots,” New Mexico’s attorney general wrote in the filing.
The news of Zuckerberg’s blessing comes after some horrific stories of minors having wildly inappropriate interactions with his company’s chatbots. In one example, a wall street journal The author is posing as a 14 year old girl it turned out A meta bot based on pro wrestler John Cena will happily engage in sexual conversation with very little pushback.
“I want you, but I need to know that you’re ready,” Cena-bot tells what she thinks is a teenage girl. After being told that the 14-year-old wanted to take the lead, the AI ​​chatbot assured him that it would “cherish your innocence” before moving on to an intense sexual role-play.
The chatbots were released in early 2024, specifically designed for romantic and sexual engagement on Zuckerberg’s orders. WSJ notes. At the time, Ravi Sinha, Meta’s head of child safety policy, wrote that “I do not believe that creating and marketing a product that creates a U18 (under 18) romantic AI for adults is appropriate or defensible,” according to court documents.
According to court documents, company employees “lobbied hard for parental controls to shut down GenAI – but GenAI leadership deferred to Mark’s decision.”
In defense of Meta, its spokesperson Andy Stone said reuters New Mexico’s allegations are false. “This is yet another example of the New Mexico Attorney General cherry-picking documents to paint a flawed and inaccurate picture,” he said.
In any case, it seems like the company has taken the lesson to heart, at least for now. Just a few days ago, Meta announced it was shutting down teens’ access to its companion chatbots entirely, at least “until (an) updated experience is ready.”
More about Zuckerberg: Meta fires thousands of VR employees as Zuckerberg’s vision fails
