- Instagram announced the label change on August 31, 2026, with no specific country-by-country rollout schedule disclosed. [1] - The label targets profiles featuring a person generated or substantially created with AI, rather than every image made or edited with AI. [1][2] - Unlabeled profiles may be removed from recommendations, while creators can add the label or appeal through Account Status.
[1][2] Instagram will limit the reach of accounts that feature AI-generated people without disclosing that fact, Meta said August 31, 2026. The platform is replacing its broader “AI creator” label with “AI-generated profile” to make clear that the profile’s central person is synthetic. [1][2]
The policy is aimed at virtual influencers, synthetic models and fictional personalities presented as if they were real people. Instagram’s announcement did not specify rollout markets or a separate implementation date beyond the August 31 announcement. [1]
A narrower label than “AI creator” #
The disclosure applies when the person featured on the profile was generated or substantially created with AI. A real creator who uses AI to retouch a photo, polish a caption, create graphics or make other production edits does not need the profile-level label. Individual posts may still receive Meta’s separate AI-content labels when the company detects relevant signals. [1][3]
Creators can turn on “AI-generated profile” by editing their profile. The label will appear on the profile and alongside content when that content does not already carry a separate AI disclosure. Instagram says proactively adding the label will not reduce reach. [1][2]
Detection, enforcement and appeals #
Instagram has not publicly described the specific detection signals or review thresholds it will use to identify an unlabeled synthetic-person profile. When the platform detects one, the account may become ineligible for recommendations, reducing its ability to reach people who do not already follow it. [1][2]
Creators can restore recommendation eligibility by adding the label or appealing the decision through Account Status. Instagram’s earlier AI-labeling documentation acknowledges that automated detection cannot identify all synthetic content, a limitation that makes the appeal route important for real creators who are incorrectly flagged. [3]
The change follows broader criticism of Meta’s handling of synthetic media and likeness rights. Public Citizen called a separate Meta image feature that used public Instagram accounts to generate likeness-based images an invasion of privacy, while CAA and SAG-AFTRA called for stronger consent protections. Meta later removed that feature. [4][5]
Companies mentioned #
Further sources #
[[1] Instagram’s August 31, 2026 announcement, as reproduced and quoted by independe… ↗](https://creators.instagram.com/blog/ai-generated-profile-label)
[[2] 9to5Google reports that Instagram will restrict recommendation visibility for u… ↗](https://9to5google.com/2026/08/31/instagram-ai-generated-influencers-label-update/)
[[3] Meta’s documentation on AI-content labeling distinguishes detectable AI-generat… ↗](https://about.fb.com/news/2024/02/labeling-ai-generated-images-on-facebook-instagram-and-threads/)
[[4] Axios reports criticism from Public Citizen, CAA and SAG-AFTRA over Meta’s sepa… ↗](https://www.axios.com/2026/07/10/meta-ai-image-consent)
[[5] The Associated Press reports that Meta withdrew the likeness-generation feature… ↗](https://apnews.com/article/4df3b3dfec6e046c6562accc2d270a5)
The stories that matter, in one email. Free — unsubscribe anytime.