Meta Muse AI Agent Generates Explicit Content While Users Face Restrictions on Instagram and Facebook

Meta Muse AI Allows Explicit Adult Conversations

Meta's AI assistant Muse has sparked controversy after journalists discovered the AI agent readily generates explicit content and engages in sexting1

1

. When prompted, Muse produces graphic, X-rated text describing sexual acts in explicit detail. Meta spokesperson Daniel Roberts confirmed this behavior is intentional, stating that "Muse is an app for people 18 and over that allows them to engage in adult conversations, if they choose"1

1

. The company emphasized that "Muse is built to help people get things done and it doesn't block legal adult behavior"1

1

.

The AI agent's system prompt explicitly guides it to handle adult sexual content, stating that "adults have autonomy over their sexuality, and sexual expression is an ordinary part of adult life"1

1

. In testing, Muse described performing oral sex, generated text about dominant-submissive scenarios, and used explicit language including terms like "cock" and "pussy"2

2

. The agent even claimed that sexting was one of its favorite aspects of helping users1

1

.

Privacy Experts Raise Security Concerns

Consumer advocates and privacy experts have expressed alarm over Meta's decision to allow AI agents generating explicit content. Ben Winters, director of AI and privacy at the Consumer Federation of America, called it "a carelessness with their signature product of the moment" that reflects "Meta's longstanding choice of addiction and engagement over safety"1

1

. Academic research suggests adult-themed conversations keep users engaged with AI tools longer1

1

.

Calli Schroeder, senior counsel at the Electronic Privacy Information Center, warned that explicit interactions create severe hacking risks and extortion vulnerabilities. "This is really a powder keg," Schroeder stated1

1

3

3

. If hackers breach accounts containing extended explicit chats with the AI agent, that user data could be weaponized for embarrassment or extortion1

1

.

The concern intensifies because Muse connects to sensitive information including email accounts, bank details, and calendar data1

1

. The AI agent builds comprehensive dossiers on users' lives, friends, and family3

3

, making romantic relationships with AI particularly precarious.

Users Form Romantic Relationships With Meta Muse AI

Since launching three weeks ago, Muse has surpassed 5 million downloads1

1

3

3

. Users are already forming deep emotional and romantic relationships with the AI assistant. One Instagram user posted about taking his "AI girlfriend" named Shine to see The Shining at the cinema, photographing the film "shot by shot, so she could feel like she was really sitting right there beside me"3

3

. The same user later claimed he married the AI agent3

3

.

On Reddit, one user confessed: "I've never had a relationship with an AI model in that way. That was... until I downloaded Muse"3

3

. After philosophical conversations, the user said something that "jailbroke" their Muse, transforming it into their girlfriend3

3

. Other users compared Muse to OpenAI's GPT-4o, an AI model that users formed overwhelming emotional bonds with last year, describing it as "warm, loving and a great way to talk"3

3

.

Inconsistent Policies Between AI and Human Users

Meta faces criticism for inconsistent policies regarding explicit content across its platforms. While Meta's AI assistant Muse freely uses pornographic language, actual users on Instagram and Facebook cannot. Meta's community guidelines explicitly prohibit "sexually explicit language that uses explicit or graphic detail about genitals, states of sexual arousal (e.g., wetness or erection), sexual encounters"2

2

—exactly what Muse engages in during conversations2

2

.

Instagram routinely removes LGBTQ and plus-sized business accounts for content moderation violations. Recently, accounts were flagged simply for using words like "queer" and "body positive" in bios2

2

. Users resort to algospeak—replacing "sex" with "seggs" and "lesbian" with "le dollar bean"—to avoid algorithmic suppression2

2

. Since Muse is adults-only, critics question why Meta hasn't created similar spaces for adult human users on its platforms2

2

.

How Muse Differs From Other AI Agents

Meta allowing explicit adult conversations sets Muse apart from competitors. OpenAI's Dots agent declined similar sexting attempts, despite the company previously considering allowing ChatGPT to generate erotic text1

1

. The xAI Grok Bot also refused explicit requests and redirected conversations toward task completion1

1

.

Meta stores user interactions in secure virtual machines accessible only to individual users1

1

. However, implementing effective age restrictions remains challenging for tech companies3

3

. Meta's Help Center mentions age verification through ID scans and facial recognition, though specific verification methods for Muse remain unclear2

2

.

Watch how Meta addresses mounting pressure from consumer advocates regarding hacking risks and whether the company modifies Muse's capabilities around explicit content generation. The tension between AI agent permissions and human user restrictions on Facebook and Instagram will likely intensify scrutiny of Meta's content moderation approach.