If you saw the panic videos last week telling you to change an Instagram setting before strangers could make AI images of your face — and then saw headlines that Meta cancelled the whole thing — you’re probably wondering whether you still need to do anything.
Short answer: yes, one 30-second check. Meta pulled the specific feature everyone was angry about. It did not pull the setting that governed it, the generator behind it, or the default that opted you in. Every guide that went viral last week was written before the reversal, so right now the internet is split between instructions for a feature that no longer exists and headlines implying you can relax entirely. Both are wrong in different ways.
Here’s the calm, current version — what actually happened, what’s still turned on for your account, and the honest part most of the viral posts got wrong.
What just happened, precisely
On July 7, Meta launched Muse Image — its first in-house AI image generator, free inside Instagram, Facebook, WhatsApp, and the Meta AI app. One promoted capability let anyone @-mention a public Instagram account in a prompt and generate AI images using that person’s photos. It was on by default for public accounts over 18, and the person whose face was used got no notification.
The backlash took about 48 hours to go industry-wide. A CNET reporter generated a fake image of a colleague in under a minute — the colleague had no idea until she was told. SAG-AFTRA told its 160,000+ members to opt out. The talent agency CAA publicly demanded Meta flip the default. Security firms published emergency how-to guides, and one creator’s walkthrough of the opt-out setting pulled close to four million views in two days.
On Friday, July 10, Meta reversed course. Its statement: the feature “missed the mark, so it’s no longer available.”
Now the part almost every headline blurred: Meta pulled the @-mention feature — not Muse Image. The generator is still live and free. The ad-tools integration is untouched. And Meta has confirmed Muse Video is still in development. What died on July 10 was one capability, not the tool and not the policy underneath it.
The 30-second check that still matters
The setting the whole controversy ran through is called Sharing and reuse, and it still exists, still defaults to on for public accounts, and still governs whether your posts can be reused with Meta’s AI features — the @-mention feature was just one thing it fed. Here’s the check:
- Open Instagram and go to your profile
- Tap the ☰ menu (three lines, top right), then Settings and activity
- Scroll to “Sharing and reuse”
- Find “Allow people to reuse your content on Instagram and with AI features at Meta”
- Toggle it off for both Posts and Reels
Two details from the fine print that the viral videos skipped:
- There’s a per-post version too. On any single post, the three-dot menu has a “Turn off reuse” option — useful if you want most content shareable but specific photos protected.
- There’s an audio toggle. The same Sharing and reuse screen includes “Original audio on Meta AI” — that’s your voice in Reels, handled separately from your photos.
If you don’t see the setting: it was still rolling out region by region as of this weekend. That’s not a sign you’re safe — it’s a sign your account hasn’t gotten the toggle yet. Check again in a few days, or use the blunt instrument below.
The option that covers everything: switch your account to private. Private accounts were excluded from the photo-reuse feature entirely, and under-18 accounts had tighter limits by default. If your account exists for friends rather than an audience, private remains the single strongest setting on Instagram — for AI reasons and every other reason.
What turning it off does NOT do
This is the paragraph the panic videos got wrong, and it’s the most important true thing in the story.
Opting out stops future use. It does not undo past use. If someone generated an AI image using your photos between July 7 and July 10 — or before you flipped the toggle — turning the setting off does not delete that image. There is no button, no request form, no mechanism to find or remove AI images other people already made. Business Insider and multiple security researchers confirmed this explicitly: existing generations remain untouched.
Two more honest limits:
- The toggle doesn’t cover AI training. “Sharing and reuse” controls whether other people can use your content with AI features. Whether Meta uses your public posts to train its models is a separate policy. In the EU and UK, privacy law gives you a formal objection form for that (Meta has said it honors previously submitted objections). In the US, no equivalent exists — the practical controls are making your account private and limiting what you post publicly.
- “Public” still means public. AI or no AI, a public account’s photos can be screenshotted, downloaded, and reused in ways no toggle prevents. That was true before July 7. The AI feature made it faster and more vivid; it didn’t invent the exposure.
None of this is a reason to panic. It’s a reason to spend the 30 seconds, understand what the toggle does and doesn’t buy you, and make the public-vs-private decision with clear eyes.
Why this will happen again
Meta pulled the feature, but everything that produced it is still in place: the generator, the default-on setting, and a product pipeline with Muse Video explicitly still coming. The image version shipped opted-in by default and lasted 72 hours under public pressure. It is not cynical to expect the video version to arrive with similar defaults — it’s just pattern recognition.
Researchers who study interface design have a name for the underlying move: defaults do the deciding. A 2026 systematic review of “dark pattern” experiments found that default-on data sharing produces dramatically higher participation than default-off, and that burying a setting several menus deep sharply reduces how many people ever change it. On-by-default, three menus deep, no notification — the Muse launch was a textbook case, which is why the reversal took a celebrity-sized backlash rather than a settings page.
The practical takeaway isn’t outrage; it’s a habit: when a Meta app announces a new AI feature, check Sharing and reuse that week. You now know where it lives and what it controls, which puts you ahead of roughly everyone who watched the panic videos and concluded the story was over.
What this means for you
If you have a public Instagram for friends and family: do the 30-second toggle check today, and seriously consider going private. You lose nothing and close the whole category of problem.
If you’re a creator or run a business account: you probably want discoverability, so going private isn’t realistic. Flip the reuse toggles off, use per-post “Turn off reuse” for anything with your face or your kids, and know that your public brand photos remain public in every non-AI sense.
If you already turned it off during the panic: you’re done — the toggle survived the feature’s removal, and your setting still stands. Just don’t assume it deleted anything created before you flipped it, because it didn’t.
If you never saw the setting at all: your region may not have it yet. Recheck within the week; until then, account privacy and what you post publicly are your levers.
If you mostly use WhatsApp or Facebook: this story was Instagram-centric, but Meta AI runs across all four apps and each one has different controls. We wrote the full four-app walkthrough here: How to turn off Meta AI on Instagram, Facebook & WhatsApp.
What this can’t fix
- Images already generated. No toggle removes them, and no removal mechanism exists.
- Screenshots and ordinary misuse of public photos. The toggle governs Meta’s AI features, not the internet.
- Model training in the US. There’s no American opt-out form for training on public posts. EU/UK residents have the objection form; everyone else has account privacy.
- The next feature’s defaults. Turning this off doesn’t pre-opt you out of whatever ships next — Muse Video included. The setting-check habit is the durable protection.
The bottom line
Meta cancelled the feature, not the setting. The 30-second check — profile → ☰ → Sharing and reuse → both toggles off — is still worth doing today, and it doesn’t delete anything already made, so the sooner the better. The people who understand what their settings actually control are consistently the calmest people in these news cycles — if you want that to be you, our AI privacy for everyday tools course covers the same clear-eyed approach across every app you use, and the first two lessons are free.
Sources
- Meta’s Muse Image launch announcement — Meta Newsroom
- Meta removes controversial AI feature on Instagram after backlash — TechCrunch
- Meta Removes Muse AI Instagram Feature After Backlash: ‘Missed the Mark’ — TheWrap
- Instagram users: Here’s how to stop Meta’s AI from using your photos — TechCrunch
- CAA Slams Meta for Using Opt-Out Policy for AI Platform Muse Image — Variety
- Instagram’s New AI Update Faces Blowback From Hollywood, Cybersecurity Companies — Forbes
- Meta Pulled Muse Image’s Most Controversial Feature. The Part That Matters Ships Anyway. — Forbes
- Turn off this Meta setting before someone generates AI images of you — Malwarebytes
