Instagram is changing how it labels AI-generated profiles and will start limiting the reach of accounts that feature AI-generated people without saying so.

The existing "AI creator" label is being renamed "AI-generated profile". The company says this should make the disclosure easier to understand. The new label tells users that the person featured on a profile was generated or substantially created with AI.

How the rule works

  • Accounts that do not label will see their reach and appearance in recommendations reduced.
  • Accounts that do label are not penalised simply for having an AI-generated person as their subject — they keep full reach.
  • Tool-level use is out of scope: editing photos, polishing captions, creating graphics or making other creative tweaks does not require the label.
  • There is a route back: affected accounts can add the label afterwards or appeal via Account Status.

The reason, in the company's own words

Instagram says the change comes partly in response to users who stumbled across profiles that looked like they belonged to real people, only to find out later that the person was completely AI-generated.

Its statement: "As generative AI becomes a bigger part of how people create, 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. They want to know when a profile features an AI-generated person."

That sentence is itself an admission: the platform is conceding that its users cannot tell AI profiles from real people.

What it follows

The timing is notable. Frustration has been growing as AI influencers became more common across social platforms.

Earlier this year the dating app Goose became the subject of a Wired investigation after a network of apparently AI-generated male influencers promoted the app on Instagram. Wired found more than two dozen accounts that appeared to feature AI influencers, some of which reached out to potential users through direct messages to get them to sign up.

Health and wellness content is a more worrying example still: in July the New York Times reported finding hundreds of AI-generated doctors, healers and wellness personalities on social platforms.

There is a limit to it: the rule looks at the person on the profile, not at the content. AI-generated videos posted from a real person's account fall outside this label. The policy targets fake identities, not fake content.

The scale problem

Some numbers put the decision in context. Twenty-one percent of YouTube Shorts are already AI-generated. Thirty-five percent of all new websites are fully or partly AI-written.

Meta's own history is instructive. In late 2024 the company was planning a "coexistence" of AI profiles and human users, and even tried launching official AI characters. That went nowhere.

Within two years the position has reversed: from offering AI profiles as a product to throttling the ones that go undisclosed. What changed in between is that someone measured what users actually wanted.