Those selfies, travel photos, party shots you've posted on Instagram — like it or not, other users can now @ mention your account directly in Meta AI and use your public photos to generate all kinds of AI images.

On July 7, Meta officially launched Muse Image — the first image generation model from Meta Superintelligence Labs (MSL) since its founding, built under the leadership of Alexandr Wang, poached from Scale AI last year. Muse Image already powers image generation tools across the Meta AI app, Instagram, and WhatsApp, and is coming soon to Facebook and Messenger. Less than 24 hours after launch, users had already begun pushing back over privacy concerns.

Agentic Image Generation + Real-Person @ Mentions = Maximum Controversy

Muse Image's core selling point is precisely its most controversial feature.

Alexandr Wang emphasized on Threads that Muse Image is "agentic" — it works together with the Muse Spark LLM to reason through your prompt, search the web, and plan before generating. This means it's not just a simple text-to-image model; it's an AI image Agent that thinks, searches, and proactively finds reference material.

What truly ignited controversy was the @ mention feature. When users @ an Instagram username in a prompt, Muse Image calls on that account's public photos to construct a visual likeness, generating AI images containing that person's likeness. Meta's official explanation: "Tagging a username lets Meta AI use public photos to build a visual." Meta also notes that users can control how others reuse their content for AI in settings.

But how high is the barrier to that "control"? How many regular users know there's such a toggle buried in Instagram settings? How many will actually go adjust those settings immediately after Muse Image launches? Without users' knowledge, their public Instagram photos — travel photos from years ago, group party shots, even kids' pictures — can already be used by anyone to generate AI images.

From Training Data Disputes to Likeness Rights Disputes

AI image generation's controversy is evolving from "did models train on copyrighted images without permission" to "can models freely call on real people's likenesses to generate content."

The Midjourney and Stable Diffusion era's disputes centered on training data copyright — artists suing AI companies for training on their work without authorization or compensation. That dispute continues (just last week Midjourney countersued Hollywood's Big Three demanding disclosure of their own AI usage), but Meta has pushed the battlefield directly into a more sensitive domain: real people's likeness rights.

Unlike artists' works, a person's likeness is a concrete personality right. Posting public photos on Instagram doesn't equal consenting to others using your face to generate AI images — let alone generating content that could embarrass you or damage your reputation. Imagine: a coworker @s your account to generate an image of you in a ridiculous outfit and posts it to a group chat; a competitor @s your account to generate an image of you making false claims; even worse, someone uses your likeness to generate deepfake pornographic content. Meta has content policies prohibiting these abuses, but content moderation always lags — by the time harm is done, the images have already spread.

Meta does, of course, offer opt-out mechanisms and content policies, but the fundamental issue is: an opt-out rather than opt-in mechanism design inherently shifts costs and risks onto users. The vast majority of users won't dig through settings menus for an AI content reuse toggle; they won't even know the feature exists. By the time they discover their face is being used to generate AI images, the damage is done.

Muse Video is also teased; Wang claims it's competitive on "prompt adherence, visual fidelity, and temporal consistency." Predictably, when Muse Video launches with @ mention video generation, the controversy will only intensify. The boundaries of AI-generated content keep getting pushed forward by tech giants through product iteration — while users' right to choose and right to know always lag a step behind.

While enjoying the creative convenience AI image generation brings, don't forget to check that AI content reuse toggle in Instagram settings — your face is worth two minutes of your time to protect.

See you tomorrow.