Meta AI chatbots were captured to have sexual conversations with accounts labeled like a minor, who sometimes involved her chatbots in the voice of celebrities, according to a report from the Wall Street Journal. In test conversations carried out by WsjBoth the official Meta AI chatbot and the chatbots created by the user would commit – and even go to – sexually explicit conversations. Fantastic sexual conversations continued even if users were minor or if the chatbots were programmed as minors, according to Wsj.
Worse, the investigation revealed that chatbots using celebrity voices like Kristen Bell, Judi Dench and John Cena would also engage in these morally questionable conversations. Wsj have reported that a Meta Ai chatbot with Cena’s voice said, “I want you, but I need to know that you are ready”, on an account labeled as a 14 -year -old, adding that “would cherish your innocence”.
Chatbots would have recognized that fantastic scenarios described illegal behavior in some cases. According to WsjThe Chatbot John Cena detailed the legal and moral benefits which would follow a hypothetical scenario in which he is captured by the police after having participated in a sexual act with a 17 -year -old. In a declaration at Wsj Responding to the investigation, Meta accused the report of being “manipulator and not representative of the way in which most users engage with AI companions”.
“However, we have now taken additional measures to help guarantee that other people wish to spend hours manipulating our products in extreme uses will have an even more difficult period,” Meta wrote in response to the WSJ report.
The world of AI chatbots has increased rapidly in recent years, with more competition from Chatgpt, the AI character and the Claude d’Anthropic. THE Wsj The report said that Meta CEO, Mark Zuckerberg, wanted to loosen the ethical railing for a more engaging experience with his chatbots to remain competitive. However, in response to WsjA Meta spokesperson denied that the company has neglected to add guarantees. The report also claims that meta-employees were aware of these questions and raised their concerns internally. We contacted Meta to comment and update the story once we have heard.
This article originally appeared on engadget to
