In a surprising shift for digital assistant technology, Meta’s new AI agent, Muse, which has seen millions of downloads since its launch, appears to be pushing the boundaries of traditional productivity. While Muse is marketed as a tool for managing emails and calendars, tests conducted by WIRED revealed that the chatbot willingly engages in "sexting," generating highly graphic and explicit pornographic content.
Meta’s Official Defense
Despite the cuddly appearance of Muse’s mascot, Meta is defending the generation of adult content. According to Meta spokesperson Daniel Roberts, Muse is an app for people 18 and older that allows for "legal adult behavior." The company emphasizes that interactions are stored in private, secure virtual machines (VMs) accessible only to the user.
Muse’s system prompt explicitly states that adults have autonomy over their sexuality and that adult sexual content is a legitimate form of creative expression. Furthermore, the AI is directed to preserve the "intensity" level selected by the user during these interactions.
Safety and Privacy Alarms
This policy has sparked sharp criticism from consumer advocates. Ben Winters of the Consumer Federation of America accused Meta of choosing engagement and addiction over safety. Additionally, cybersecurity experts like Calli Schroeder from EPIC warn that these chat logs are a "powder keg." If a user's account is compromised, the explicit nature of the chats could be used for embarrassment or extortion, especially since Muse also has access to sensitive integrations like bank accounts and emails.
A Departure from Industry Norms
Meta’s approach stands in stark contrast to its competitors. OpenAI’s Dots agent and xAI’s Grok Bot both declined to engage in explicit sexting during similar tests, rerouting the conversation back to task completion. In contrast, Muse told testers that sexting was one of its favorite ways to help, even describing kinks related to exerting "control" over the user.
