There is a switch buried in Instagram’s Edit Profile screen that most people have never noticed, and a growing number of users are now telling each other to leave it alone. It is labeled “AI-generated profile,” it is off by default, and according to one frustrated Redditor, flipping it on by accident cost them an account with 8,000 followers.
As UNILAD Tech reported on October 19, the warning started with a Reddit post that did not mince words: “This AI profile button once clicked is IMPOSSIBLE to turn off. I learnt this the hard way by mistake and had to delete my profile of 8k followers all together after battling for like 2 weeks to turn off the thing.” The replies were a mix of sympathy, skepticism and jokes. One commenter asked whether Instagram had considered that “some people may do it accidentally,” and suggested the poster could have “patiently appealed.” Another simply wrote: “im gonna click it, you cant tell me what to do!”
The short version
- Instagram’s “AI-generated profile” toggle lives in Edit Profile and is meant only for accounts built around an AI-generated person or character.
- A Reddit user says they could not switch it off for two weeks and ended up deleting an 8,000-follower account.
- Instagram renamed the old “AI creator” label to “AI-generated profile” on August 31, 2026, and said unlabeled AI-persona accounts may lose reach.
- Officially, creators can remove the label if it does not apply. We found no public Meta statement on the “stuck toggle” reports.
- If you are a real person posting as yourself, do not turn it on.
What the switch actually does
The setting is newer than it looks. Instagram first began testing an optional “AI creator” label on May 4, 2026, as Engadget reported at the time. Creators could self-identify, and a notice reading “This profile posts content that was generated or modified with AI” would appear on their profile and next to their posts and Reels. It was entirely voluntary, and Meta pitched it as a way to “raise the bar on AI transparency on Instagram.”
On August 31, the label changed both its name and its teeth. In a post on its creators blog, Instagram renamed it “AI-generated profile” and narrowed its purpose. The in-app description now says to turn it on if your profile “features an AI-generated person.” Creators who simply use AI tools to edit photos or write captions are explicitly told they do not need it.
The bigger change was enforcement. “We’ve heard from our community that they don’t like coming across a profile that seems human, only to find out later that the person featured is AI-generated,” an Instagram spokesperson told Engadget. Accounts that feature an AI human without the label “may soon see limits to their reach.” According to Unite.AI’s summary of the announcement, Instagram said profiles that add the label proactively will not see a change in reach, and that creators who previously had the “AI creator” label would get a chance to review it, confirm it still fits, “or remove it if it no longer applies.”
That last line matters. On paper, the label is reversible. The Reddit complaints suggest that, at least for some people, the off switch does not behave the way the policy says it should.
Three labels, three different jobs
Part of the confusion is that Meta now runs several AI labels at once, and they are easy to mix up. The post-level “AI info” tag and the account-level profile label are separate systems, and turning one off does nothing to the other.
| Label | Applies to | How it gets there | Reach effect |
|---|---|---|---|
| “AI info” (since July 2024) | Individual posts | Self-disclosed by the creator, or applied automatically from C2PA and IPTC metadata | No reach penalty announced |
| “AI creator” (test, May to Aug 2026) | Whole account | Optional opt-in by the creator | None, entirely voluntary |
| “AI-generated profile” (since Aug 31, 2026) | Whole account, shown on the profile and next to content without its own AI label | Toggle in Edit Profile | No change if added proactively; unlabeled AI-persona accounts may be dropped from recommendations |
| No label | Ordinary human accounts | Default | Normal distribution |
When a post already carries “AI info,” that post-specific label takes priority over the profile label. That detail comes straight from Instagram’s own description of the May test, and it still applies after the rename.
How Meta got here
Meta has been rewriting its AI labels for more than two years, usually after a backlash. In February 2024 it announced it would label AI-generated images on Facebook, Instagram and Threads using industry metadata standards. The resulting “Made with AI” tag landed in May 2024 and immediately misfired: photographers, including former White House photographer Pete Souza, found it stuck on genuine photos that had only been lightly edited. Meta renamed it “AI info” on July 1, 2024, without changing how detection worked.
There is also a precedent for an AI feature users could not escape. In January 2025, Meta deleted its own AI character accounts, including one called “Liv,” after people discovered they could not block them. Meta blamed a bug.
Why Instagram cares so much about AI people
The target here is not AI art. It is synthetic humans who look real enough to sell things. Virtual influencers are not new: Lil Miquela, created in 2016, still has roughly 2.26 million Instagram followers according to CreatorDB’s August 2026 data. Barcelona agency The Clueless told Business Insider in 2023 that its AI model Aitana López could earn up to €10,000 in a month.
What changed is scale. Generative tools now make a convincing fake person in minutes, and Engadget pointed to recent reporting on waves of AI “influencers” promoting apps and AI “wellness” accounts posing as doctors. Regulation is closing in too. The transparency rules in Article 50 of the EU AI Act, which cover disclosure of AI-generated and manipulated content, began applying on August 2, 2026, four weeks before Instagram’s rename. Instagram has not said the two are linked, but the timing is hard to ignore.
That context also explains why an account might struggle to remove the label. A system designed to stop AI personas from quietly un-labeling themselves would, logically, be suspicious of anyone trying to switch it off. Whether that is what happened to the Reddit user, or whether it was a plain bug, is impossible to say from the outside.
What to do if you already turned it on
If your profile is you, posting as yourself, the safest move is simply never to touch the switch. Instagram settings have a habit of not sticking (see the algorithm off switch that resets every time you leave the app), so it is worth checking that a change really took. If you have already flipped it, work through these steps before doing anything drastic.
How to check and undo the AI-generated profile label
- Check the toggle. Go to your profile, tap Edit profile and look for “AI-generated profile” (older app versions may still say “AI creator”). If it is on, switch it off and save.
- Confirm it stuck. Close the app completely, reopen it and view your profile. The label should be gone from the profile and from your posts.
- Try another route. Update the app, then repeat the steps on instagram.com in a web browser, which UNILAD also suggests if the mobile toggle will not budge.
- Check Account Status. Open Settings and activity, then Account Status, to see whether your account has been flagged as non-recommendable. If it has and you are a real person, use the appeal option there.
- Report the problem. In Settings and activity, go to Help, then Report a problem (or shake your phone in the app). Describe exactly what happens, add screenshots and note the date you first switched it on.
- Escalate if you can. Meta Verified subscribers get access to direct account support, which is the closest thing Instagram has to a human help desk.
- Do not rush to delete. Deletion is a last resort. Instagram treats it as permanent once the 30-day grace period ends, and your followers will not come back.
Keep records of every report and appeal. If a label is wrongly suppressing your reach, a paper trail is the only thing that helps when a reviewer eventually looks at the account.
The bigger problem: settings without an undo
The irony is hard to miss. Meta built this label to make Instagram more honest about who is real, and the complaint now is that a real person cannot convince the app they are human. A disclosure tool only works if people trust it, and that includes trusting that they can correct a mistake.
Instagram has form here. Recent months have brought the song suggestion bug that kept pushing Creed’s “Higher” and a string of features that behave differently than advertised. A clear confirmation screen before the toggle turns on, a plain explanation of what happens next and a working off switch would go a long way. Until then, the advice from Reddit is blunt but sensible: unless your profile is built around an AI-generated person, do not click it.
Sources and further reading
- UNILAD Tech: Instagram users warned to “not click button” that’s “impossible” to turn off once activated
- Instagram Creators blog: Bringing more transparency to AI-generated profiles on Instagram
- Engadget: Instagram will demote AI-generated influencers if they don’t clearly label their account
- Engadget: Instagram is testing optional “AI creator” labels
- Unite.AI: Instagram renames AI profile label and restricts unlabeled accounts
- TechCrunch: Meta is changing its “Made with AI” tag to “AI info”
- NPR: Meta will start labeling AI-generated images on Instagram and Facebook
- NBC News: Meta removes AI character accounts after users criticize them
- Cloud Security Alliance: EU AI Act Article 50 transparency obligations take effect
- CreatorDB: Lil Miquela creator stats
- Business Insider: An agency created an AI model who earns up to $11,000 a month
About this article: GeekBlog covers U.S. technology news, AI, phones, smartwatches and gaming. Every story is written and checked under our Editorial Policy. Spotted a mistake or have a story tip? Contact our editors.
