How to Protect Your Privacy from Instagram’s Muse AI: Opt-Out Guide and What to Know

Earlier this month, Meta launched a new AI image tool on Instagram called Muse. Within days, it drew backlash from Hollywood, cybersecurity firms, and the actors’ union SAG-AFTRA, which urged members to opt out. By July 10, Meta removed the feature, admitting it “missed the mark.” But the underlying privacy questions—about how your photos can be used to train generative AI—remain live for anyone posting on Instagram.

Here’s what the Muse episode told us, and what you can do now to protect your likeness and data.

What Happened

Muse was an AI image generator that used your own Instagram photos as a base. According to reports from Yahoo Tech (July 10, 2026), the tool could reimagine you in different artistic styles—think turning a selfie into a charcoal sketch or a Pop Art poster. Meta positioned it as a creative feature, but critics pointed out that it effectively let the company train AI models on user images with little upfront transparency.

The backlash was swift. Yahoo reported that cybersecurity companies flagged the privacy risks, noting that users’ faces could be used to generate new images without explicit consent for each use. SAG-AFTRA specifically recommended members opt out of Meta’s AI training, calling it necessary to “take action to protect your likeness.” By July 10, Meta pulled Muse, saying in a statement that the feature did not meet user expectations.

Why It Matters

Even though Muse is gone, the underlying AI training pipelines remain. Meta has tools that learn from public posts, and the company can reintroduce similar features later. The core risk is “likeness theft”: an AI could mimic your appearance or style in a context you never approved, such as in a fake advertisement, a deepfake video, or a misleading post.

For creators and influencers who regularly share high-quality images of themselves, the stakes are higher. A recognizable face is a brand asset. For everyday users, the concern is less dramatic but still real: your photo could appear in synthetic images you never made, used in ways you can’t control. Once a likeness is fed into a generative model, it’s hard to remove completely.

What Readers Can Do

Even without an active Muse, you should review your Instagram settings to limit how Meta uses your data for AI training. Here’s a step-by-step guide:

  1. Open Instagram Settings – Go to your profile, tap the hamburger menu (top right), then tap “Settings and privacy.”
  2. Find Data Sharing – Scroll down to the “Account” section and tap “Data sharing” (the exact label may vary by region and app version).
  3. Locate AI Training Controls – Look for a toggle or link labeled “Allow your data to be used for AI training” or something similar. If you see it, switch it off. If you don’t see it, check under “Privacy” or “Data policy.”
  4. Opt Out Globally – You can also visit Meta’s Privacy Center (accessible from settings) and look for “Generative AI” controls. If no toggle exists in your country, you may need to submit a request through Meta’s help forms.

Beyond settings, adopt these habits:

  • Watermark your images – Use visible or invisible watermarks on photos you post publicly. Even a subtle overlay makes it harder for scrapers to claim the image is entirely yours.
  • Limit your audience – Switch your account to private, or use “Close Friends” for sharing personal photos. Public posts are far more likely to be used for training.
  • Review your history – If you have older public photos you’re uncomfortable with, consider deleting or archiving them.
  • Monitor for misuse – Periodically search for reverse-image matches of your photos or use services like Google Images search. If you spot a synthetic image that looks like you, report it to the platform.

Finally, follow SAG-AFTRA’s advice: if you’re a professional, treat your image as a controlled asset. For anyone using Instagram, assume that public photos can be used in AI training unless you proactively opt out.

Sources

  • Yahoo Tech. “Meta Removes Muse Image AI Feature After Backlash: ‘Missed The Mark’.” July 10, 2026.
  • Yahoo. “SAG-AFTRA Recommends Members Opt-Out Of Meta’s AI Feature: ‘Take Action To Protect Your Likeness’.” July 9, 2026.
  • Yahoo Finance. “What Meta’s Muse AI Image Tool Means for Instagram Privacy.” July 10, 2026.