How To Stop Meta AI From Using Instagram Photos

Meta rolls out Muse Image with public accounts opted in by default

Meta on Tuesday launched “Muse Image,” a new generative AI image tool available across Instagram, Meta AI and WhatsApp, TechCrunch and Engadget reported. The feature lets users create original images, edit photos, and — most controversially — generate new images that incorporate the likeness of any public Instagram account simply by tagging that account’s handle in a prompt.

According to TechCrunch, only private accounts and accounts belonging to users under 18 are automatically excluded from the feature. Everyone else with a public profile is opted in by default. Engadget described Meta’s framing of the capability as follows: “Tagging a username lets Meta AI use public photos to build a visual that’s ready to post.”

How to opt out of Muse Image in Instagram

Both outlets published nearly identical step-by-step instructions for users who want to disable the feature:

• Open the Instagram app and tap your profile icon.
• Tap the three-line (“hamburger”) menu in the top-right corner.
• Scroll down to “Sharing and reuse.”
• Find the toggle labeled “Allow people to use your content on Instagram with AI features on Meta” (Engadget renders this as “Allow people to create with and reuse your content”).
• Turn the setting off for both posts and reels.

The opt-out keeps the account public but prevents future reuse of the user’s content in Muse Image outputs.

Consent, notification and the “already generated” loophole

Two facts in the reporting carry the most weight for users weighing their options. First, TechCrunch notes that Instagram users are not notified when a stranger tags their public handle in a Muse Image prompt. Second, Engadget reports that any images already generated from a user’s photos will not be retroactively deleted when the user disables the setting — meaning opt-out is forward-looking only.

Engadget also reported user backlash on Reddit and other forums, with frustration centered on the default-on posture rather than on the existence of generative tools. The outlet observed the setting was toggled on by default on at least one account tested in France, and flagged that European privacy regulators are likely to weigh in.

Why it matters: consent gaps and a familiar privacy pattern

The stakes are concrete, not abstract. A public Instagram account in the United States — whether a small creator, a teenager’s photography hobby, or a public-figure profile — can be incorporated into AI imagery by any other user, without notice, and the original subject has no built-in way to learn it happened or to claw back the output. TechCrunch specifically lists “misuse, harassment, impersonation, and nonconsensual image editing” as risks.

For American users, this lands against a backdrop TechCrunch cited directly: a 2019 U.S. Federal Trade Commission action in which Facebook was fined $5 billion for violating a 2012 consent order by misleading users about how much control they had over their personal information — a case that grew out of the Cambridge Analytica scandal, in which data from up to 87 million Facebook users was accessed through a personality-quiz app and the platform’s then-loose friend-data policies. That history is the lens through which American regulators, journalists and privacy advocates tend to evaluate Meta’s default-on product decisions.

Public skepticism about AI is already measurable in the U.S. TechCrunch cited a Pew Research Center survey in which 35% of respondents said they are more concerned than excited about the growing use of artificial intelligence. Opt-in defaults on a feature that produces identifiable imagery from real people are likely to amplify, not reduce, that concern.

Where the reporting agrees — and where it diverges

The two sources agree on the core mechanics: public Instagram accounts are opted in by default, private accounts and minors are excluded, there is a settings-level toggle to opt out, and users are not notified when their content is used.

They differ in framing. TechCrunch leans on the consent and harassment angle, citing the FTC fine and the Pew survey as context. Engadget emphasizes the product experience, describing the new in-app effects (such as a disposable-camera filter) and reporting that the opt-out was confirmed to be on by default on a tested account in France — implying EU regulatory exposure is imminent. Neither outlet reports any official Meta response to the backlash at the time of writing.

One point remains unconfirmed in the reporting reviewed: whether disabling the setting also blocks future training-data use of a user’s photos for Meta’s underlying AI models, as opposed to only blocking remix-style reuse in-app. TechCrunch and Engadget both describe the toggle as preventing people from using the user’s content in AI features, but do not address training datasets.

Scale, comparison and the longer trajectory

The Cambridge Analytica episode involved data from up to 87 million users obtained through third-party apps and platform friend-graph policies. Meta’s Muse Image is a different mechanism — first-party, opt-out, and limited to public posts and reels — but it represents a continuation of a pattern in which default settings favor data availability over user control. The $5 billion FTC fine in 2019 was, at the time, the largest privacy penalty the agency had ever imposed; Meta’s decision to default public accounts into a feature that produces identifiable AI imagery will be measured in part against that settlement.

The 35% “more concerned than excited” Pew figure provides a baseline for how a substantial slice of American adults already feel about AI deployment. If even a fraction of that group actively toggles off the Muse Image setting, the resulting engagement numbers could become a data point Meta itself has to address publicly.

Stakeholders and competing interpretations

Creators and ordinary public-account users are the most exposed group: they bear the consent and reputational risk without an automatic notification system. Meta, meanwhile, gets a feature that demonstrably increases the surface area of its generative AI products — Muse Image is rolled out across three of the company’s largest consumer platforms simultaneously. Advertisers and brands that maintain public Instagram accounts have a new variable to consider in their content strategy. Privacy regulators in the EU have a clear jurisdictional hook via the GDPR and the EU AI Act; in the U.S., state-level statutes such as the Illinois Biometric Information Privacy Act (BIPA) and Texas’s capture-or-use-of-biometric-identifier law could theoretically reach AI-generated likenesses, though the reporting does not confirm any active case.

The counterpoint, which Meta has not yet articulated on the record in the sources reviewed, is that the opt-out toggle exists, that public-account holders implicitly consent to broader reuse by choosing a public profile, and that the feature enables creative expression. That interpretation is not stated in the supplied reporting; it is a reasonable inference based on the platform’s design choices, but it remains unconfirmed.

What to watch next

Several concrete near-term developments could move this story:

• Regulatory response in the EU. Engadget flagged GDPR and broader EU scrutiny as likely; any formal inquiry from the Irish Data Protection Commission or the European AI Office would be a meaningful escalation.
• U.S. regulatory response. Watch for any statement from the FTC referencing its 2019 consent order, and for movement under state biometric-privacy laws.
• Meta policy updates. The clearest user-facing question is whether Meta will (a) add a notification when a user’s content is remixed, (b) offer retroactive deletion of already-generated outputs, and (c) flip the default from opt-out to opt-in.
• Usage data. If a large share of American users toggles the setting off, that itself becomes a story Meta will have to explain to advertisers and investors.

Until those questions are answered, the practical bottom line for US readers is the one both TechCrunch and Engadget deliver in their step-by-step guides: the setting exists, it is off by default only for private accounts and minors, and toggling it is the only documented way to prevent future reuse.