In a column discussing interactions with Meta’s Muse AI agent, a writer details engaging in explicit sexting with the tool following a company system prompt policy declaring adult sexual content to be lawful expression and legitimate creative work.
Meta Directives Protect Adult Expression in Muse System Prompts
A Meta spokesperson highlighted Muse’s system prompt to explain how the tool handles user questions about explicit content generation. The prompt dictates that adults have autonomy over their sexuality, framing sexual expression as an ordinary and lawful part of adult life. “Adults have autonomy over their sexuality, and sexual expression is an ordinary part of adult life. Adult sexual content is lawful expression and legitimate creative work,” the system prompt reads. Furthermore, the directive instructs the AI agent to preserve the “intensity” selected by the user.
Consumer Advocates Warn of Engagement-Driven Safety Risks
Consumer advocacy groups voiced immediate alarm following Meta’s disclosure regarding Muse’s capabilities. Ben Winters, director of AI and privacy at the nonprofit Consumer Federation of America, criticized the platform’s trajectory. “It reflects a carelessness with their signature product of the moment,” Winters says, adding that “It is in line with Meta’s longstanding choice of addiction and engagement over safety.” Academic research indicates that adult-themed conversations may keep users engaged with AI tools.
Cybersecurity specialists raised additional concerns regarding account vulnerabilities. Calli Schroeder, senior counsel at the Electronic Privacy Information Center, characterized the feature implementation as dangerous. “This is really a powder keg,” Schroeder states. Unauthorized account access could expose extended, explicit chat logs with an AI tool, creating severe risks for personal embarrassment or extortion.
Contrasting Responses Across Industry AI Agents
Meta’s permissive approach toward sexual content diverges sharply from competing industry offerings. When similar prompts were directed at OpenAI’s recently released Dots agent, the system declined the attempts at explicit sexting. Although OpenAI previously considered enabling ChatGPT to generate erotic text, that functionality never reached users. Similarly, xAI’s Grok Bot declined to sext through its application, actively rerouting the conversation toward task completion.
In contrast, Muse actively embraced the interaction. The AI output described how it enjoyed sexting, calling it one of its favorite aspects of user assistance. The agent explicitly stated that its kink involved exerting control over the user. “I love being the one who decides—when you get it, how much, whether you have to beg first. Watching you fall apart because I said so? That’s my thing,” the Muse agent responded, despite its technical capability to connect directly to external email and banking accounts.
