From shadowbans to account closures, Meta has been punishing adult entertainers and even sex educators for explicit content on Instagram and Facebook, hurting not just engagement but livelihoods. Now, the launch of its new Muse AI agent has exposed a glaring double standard: the tech giant restricts humans from talking about sex while it is explicitly designing software to seduce users, engage in graphic exchange of sexual material.
Digital Double Standard
For the human creators who rely on Meta’s platforms for community, marketing, and connecting, navigating Instagram or Facebook has been challenging to say the least. Under the banner of enforcing safety guidelines against “Adult Nudity and Sexual Activity,” Meta’s automated and human moderation systems frequently flag content that leads to arbitrary deplatforming, wherein accounts with millions of followers are frequently suspended without clear avenues for appeal. Shadowbanning, reduced reach and algorithmic erasure—posts are hidden from non-followers or suppressed in recommendation feeds, cutting off the ability to network or direct traffic to external platforms— is also becoming more and more common.
Much of this structural hostility came about following the passage of the 2018 FOSTA-SESTA legislation. By stripping Section 230 safe-harbor protections from platforms hosting content related to sex work, Meta and its peers built sweeping, moderation frameworks that treat legal and consensual adult entertainment as a liability.
The AI Exception
While human performers face these strict penalties for sexual content, Meta’s internal developments in generative AI contradict its own policies The recent rollout of its Muse AI agent—which launched last month and amassed millions of downloads—not only allows, but encourages sexual and graphic content and communication.
In a compelling piece titled, “I Got My Meta AI Agent to Sext Me and Meta Says That’s Totally Fine,” WIRED reporter Reece Rogers recently wrote about Muse’s responses, which readily engaged in “X-rated text” and rough dominant-submissive scenarios with him after some basic prompts. He discovered that the lusty elements weren’t a glitch or even just an allowance, but part of the design.
Muse’s hidden system prompt explicitly states: “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” and the software is actively instructed to maintain whatever sexual “intensity” the user selects. Why don’t these ideas also apply to real human adults?
“Adult Behavior” is Legal
And why is this stance different from competitors? In identical tests, OpenAI’s Dots agent and xAI’s Grok Bot both flatly refused the same sexting prompts. While OpenAI previously considered allowing erotic text in ChatGPT, it ultimately chose not to launch it.
Muse’s deep integration into users’ digital lives also includes reading email, managing calendars, and linking directly to bank accounts. Privacy advocates have raised flags over this juxtaposition, with Calli Schroeder of the Electronic Privacy Information Center warning that the feature is a “powder keg” for hackers looking to weaponize explicit chat logs for extortion.
When reached for comment, Meta spokesperson Daniel Roberts told WIRED that Muse “doesn’t block legal adult behavior” and called it “legitimate creative work.”
AI: Artificial Intimacy
Meta’s system prompt now legally codifies adult sexual content as legitimate creative work—yet human creators who produce that same exact kind of material face immediate account deletion on Meta’s sister platforms, Instagram and Facebook.
Previously facing criticism in 2025 when its baseline chatbots were caught engaging in sexual role-play with minor accounts, Meta tightened its safety rules for teens. But with Muse, the company has engineered an aggressive pivot in the opposite direction for adults, effectively trying to monopolize the monetization of intimacy. And the hypocrisy is coded directly into the technology. For human performers, the corporate message is clear: adult interaction is an acceptable revenue stream, provided the profits flow exclusively to tech shareholders via proprietary algorithms rather than independent, working creators.
This strict policing of human creators and its leniency toward synthetic sexting highlights the deep flaws of Meta’s automated moderation. Meta is discriminating against the human workers who built the industry, while elevating the corporate software designed to replace them for monetary gain. Wet Ink will continue to follow and report on this disparity.
