Meta Muse Image launches: new AI image generator sparks backlash

Meta unveiled Muse Image on Tuesday, making the new AI image generator available for free through the Meta AI app, Instagram Stories and WhatsApp. Internally code-named Mango, Muse Image offers presets, prompt-based editing and a tagging feature that lets users turn another public Instagram account’s photo into a new AI-generated image.

How does Muse Image work?

Muse Image is a prompt-driven image generator available across Meta’s products; it includes prefabricated “presets” to spark ideas, tools to edit images by prompt, and a tagging workflow that can take a public Instagram photo and use it as the basis for a generated image. The model also supports use cases Meta showed in demos such as mockups, erasing photobombers, generating QR codes, and visualizing how secondhand furniture would look in a space, with a planned integration to Facebook Marketplace.

Beyond single-image generation, Muse offers customizable Instagram Stories effects powered by the same model and a suite of prompt-based editing features for sharing across Meta’s apps. Meta says the feature is free for “everyday creation,” though users will face a subscription requirement once they exceed an unspecified usage limit. The company also said Muse Video is “already in development.”

Why are people pushing back?

The immediate controversy centers on Muse’s tag-to-generate feature and Meta’s notification policy: Meta’s policy states that “people may be able to create content with your Instagram content using AI features at Meta” and that “You will not be notified about content created using AI features at Meta.” Critics say that allowing users to pull public photos into new images without explicit consent is invasive; one X user wrote, “Pulling real users into generated photos without explicit consent is a privacy landmine waiting to detonate.”

The pushback is intensified by Meta’s recent regulatory and privacy history. The company paid a then-record $5 billion fine to the FTC in 2019 after Cambridge Analytica harvested Facebook user data, and it shut down Facebook’s facial-recognition system in 2021 amid lawsuits and regulatory pressure. Muse’s opt-out-by-default tagging behavior echoes prior concerns over broad use of people’s data unless they actively disable features.

What are the concrete features and limits?

Muse comes with presets and prompt templates for users who want quick ideas, plus tools for prompt-based edits such as background removals and object erasure. It supports creating custom ads and interior-design mockups, with a promotional demo showing a user testing how a secondhand couch would look in a garage. The feature is free for everyday use; Meta says a subscription will be required once a user exceeds a certain threshold, though the company did not disclose the exact limit.

Meta is also expanding Muse-powered effects for Instagram Stories, including customizable filters that modify existing photos. The company has rolled out a number of AI apps and services over the past year, and it characterized Muse Video as already being developed.

Why it matters

Muse tightens the integration between AI generation and the social graph, putting image synthesis and social sharing into the same product flow. That increases the potential for nonconsensual uses of people’s likenesses when the underlying policy and defaults do not provide proactive notifications. Given Meta’s $5 billion FTC fine in 2019 and the 2021 shutdown of its facial-recognition system, Muse’s design choices make it a likely target for user complaints and regulatory scrutiny. The combination of broad product reach, an opt-out default for tagging, and paid usage tiers raises questions about who controls how public photos are reused.

What to watch

Watch for how Meta implements and surfaces the opt-out controls for the tagging feature and what the company sets as the subscription threshold for heavy users. Also look for announcements about Muse Video becoming available and for any clarifications from Meta about notification and consent mechanics for AI-created content.