Glonce

Instagram admits users can't tell AI profiles from real people

productsThe Glonce editorsThe Decoder🔥69Glonce rating69Impact60Novelty65Relevance85Surprise55Players85Credibility85
Instagram admits users can't tell AI profiles from real people

Instagram is swapping its existing "AI creator" tag for a new "AI-generated profile" label, the company announced, after acknowledging that users often cannot distinguish AI-generated profiles from real people. Meta explained the reasoning directly: "We've heard that people don't like seeing a profile that seems human, only to find out later that the person featured is AI-generated." Under the new policy, profiles that lack the AI-generated label will have their reach and recommendations throttled. Affected accounts can either add the label themselves or appeal the classification through Account Status. The policy carves out an exception for creators who merely use AI as a production tool rather than presenting an AI persona: they do not need the label, and any creator who voluntarily opts into it keeps full reach. The announcement comes against a backdrop the article lays out in detail: 21 percent of YouTube Shorts are already AI-generated, Instagram and Facebook are likely just as bad or worse (though no figure is given for either platform specifically), and 35 percent of all new websites are now fully or partly AI-written. The piece also notes that as recently as late 2024, Meta was still planning a "coexistence" of AI profiles and human users on its platforms and even tried launching official AI-generated characters, an effort that went nowhere.

Key facts

  • Instagram is replacing its "AI creator" tag with a new "AI-generated profile" label after finding users often cannot tell AI profiles from real people.
  • Profiles without the new label will have their reach and recommendations throttled; affected accounts can add the label or appeal via Account Status.
  • Creators who use AI only as a production tool do not need the label; those who voluntarily add it keep full reach.
  • 21 percent of YouTube Shorts are already AI-generated, and 35 percent of all new websites are fully or partly AI-written, per the article.
  • In late 2024, Meta was still planning a "coexistence" of AI and human profiles and tried launching official AI characters, which went nowhere.

Why it matters

This is Meta effectively conceding that its earlier approach to labeling AI-generated accounts failed: the old "AI creator" tag did not stop people from being fooled into thinking an AI-generated profile was a real person. It also marks a reversal from Meta's late-2024 stance, when the company was still planning for AI profiles and humans to coexist on its platforms and even tried rolling out its own official AI characters, an effort the article says went nowhere.

Who it affects

Instagram creators who run or resemble AI-generated personas are the direct target: they now face a choice between labeling their profile as AI-generated or having their reach and recommendations throttled. Creators who simply use AI tools in production, without presenting an AI persona, are exempt and do not need to label anything. Ordinary users benefit indirectly, since the new label is meant to make it clearer when a profile they encounter is not a real person.

How to use it

Accounts that are AI-generated need to add the new "AI-generated profile" label to avoid throttling; if a profile is flagged as AI-generated in error, the owner can appeal through Instagram's Account Status tool. Creators who only use AI as part of their workflow, rather than presenting an AI character, are not required to do anything. Creators can also choose to add the label voluntarily and keep full reach, giving them an incentive to be upfront about AI use rather than avoid the label.

How solid is it

The source is a direct company announcement, and the article quotes Meta's own stated rationale verbatim, which makes the core facts, the policy change, the throttling mechanism and the appeal path, solidly sourced. The supporting statistics on AI content prevalence (YouTube Shorts, new websites) come from the article itself rather than from Meta, and no comparable hard figure is given for Instagram or Facebook specifically; the claim that those two platforms are "likely" just as bad or worse is explicitly framed as an estimate, not a measured figure.

Risks and caveats

No rollout date or timeline is given for when the new label takes effect, and no detail is provided on how Instagram actually detects or verifies that a profile is AI-generated, or on how long the Account Status appeal process takes. The comparison figures for AI-generated content are drawn from YouTube Shorts and the broader web, not from Instagram or Facebook directly, so the scale of the problem on Meta's own platforms remains an assumption rather than a documented number.

“We've heard that people don't like seeing a profile that seems human, only to find out later that the person featured is AI-generated,”

— Meta