Meta's AI Info Label on Ads: What It Means for Architecture Firms
Meta's AI info label flags AI-touched ads on Facebook and Instagram. See what triggers it, where it shows, and what changed for studios in August 2026.
In this article
If your studio plans to boost a project Reel or a rendering carousel on Instagram or Facebook before the end of 2026, Meta's AI info label may already be showing up on that ad without you asking for it. As of September 2026, Meta flags ad creative made with its own generative tools or with third-party AI software, and Instagram now runs a separate rule for accounts built around an AI-generated person. This guide covers what actually triggers the ad label, where it shows up, what counts as a small edit that skips it, and why a fictional character inside a project film is a different situation from the new account-level rule.
Key takeaways
- Meta's "AI info" label appears on the "About this ad" screen, and sometimes right next to "Sponsored," when an ad is created or significantly edited with generative AI.
- The label already covers third-party AI tools, not only Meta's own Background Generation and Image Generation features, as reported in July 2026.
- Plain resizing and color correction don't trigger the label; a photorealistic AI person in an ad can push it into the feed itself.
- Instagram's August 31, 2026 "AI-generated profile" rule targets accounts built around an AI person, not background figures inside a project film.
- A short pre-boost check of your ad creative takes a few minutes and avoids a surprise label on a paid post.
What triggers Meta's "AI info" label on an ad
Meta's AI info label is not a future rule. The marketing site Common Thread Collective reported on July 13, 2026 that the policy was "in effect today" ("Meta's New AI Ad Labels," Common Thread Collective), and Meta's own help center confirms it is active now: an ad gets the label when it was "created or significantly edited using Meta's generative AI creative features or created or edited using third-party AI tools" (Meta Help Center). Inside Ads Manager, the built-in features that count include Background Generation, Image Generation and Add Animation — tools some small firms already use to swap a sky, extend a background, or add subtle motion to a still rendering before boosting it.
For an architecture studio, the practical trigger is broader than Meta's own toolset. Content built outside Ads Manager — in an AI photo editor, an AI rendering tool, or an AI video generator such as Arch2Video — can still pick up the label once it's uploaded as an ad.
Where the label shows up — and why a photoreal AI person changes that
Most of the time, the label sits somewhere a viewer has to look for it: "the About this ad screen in the three-dot menu in the top right-hand corner of the ad" (Meta Help Center). Someone scrolling past your boosted project Reel won't see it unless they tap through.
That changes for one specific case. Common Thread Collective's July 2026 review found that when an ad features a photorealistic AI-generated person, the label can appear "directly next to the Sponsored tag in the feed itself" — visible to everyone, not tucked into a menu. That detail matters for a studio boosting a project film made with an AI video tool: a photorealistic AI person in frame is a candidate for that more visible placement, not just the three-dot menu.
One category has required disclosure for longer. Meta's help center notes that "for ads about social issues, elections or politics, advertisers are already required to disclose if the image, video or audio are created or edited with AI." A studio boosting a project film isn't in that category, but the broader AI info label extends the same disclosure logic to ordinary commercial ads.
What counts as a minor edit, and why third-party tools still count
Not everything gets flagged. Meta's help center states plainly: "Not all ads will have AI info, like when there are minor changes or enhancements like image resizing or color correction." A straight photograph of a finished building, cropped and color-corrected for a Reels ad, stays in that exempt category.
Past a resize or a color grade, the third-party rule matters more than most firms expect. Meta doesn't only watch its own Ads Manager features — it reads industry-standard signals, including C2PA Content Credentials that many AI tools now embed in exported files, and applies the label when it finds them (Meta Help Center; Common Thread Collective). That means AI-generated entourage dropped into a rendering, an AI-extended background, or an AI-generated project film built from your renderings can all carry the label once boosted, even when the edit happened somewhere other than Meta's own ad tools. If your studio already tracks Content Credentials in an AI video workflow, this is the same signal showing up again, this time inside Ads Manager.
Two Meta AI labels, two different jobs
Two labels are live at the same time as of September 2026, and mixing them up is easy. One covers a single ad's creative; the other covers an entire account.
| Label | Covers | What happens if you skip it |
|---|---|---|
| AI info | One ad's image, video or audio | Shows in the ad's menu, or next to "Sponsored" for a photoreal AI person; political ads must disclose |
| AI-generated profile | An account built around an AI person | Reduced reach for the profile if the AI person isn't labeled |
The rest of this guide focuses on why that second row doesn't describe a typical architecture studio's Instagram account, even one that posts AI-generated project films regularly.
Why a fictional cast in a project film isn't an "AI-generated profile"
Instagram's August 31, 2026 announcement renamed its "AI creator" label to "AI-generated profile" and tightened the rule around it. TechCrunch reported that accounts "featuring AI-generated people must apply it or face reduced reach" ("Instagram puts new limits on undisclosed AI profiles," TechCrunch), and Social Media Today reported the same day that the label is required specifically when "AI-generated people are the profile's main subject," with unlabeled AI-person profiles seeing reduced reach while labeled ones don't ("Instagram updates tags for AI profiles," Social Media Today).
That threshold — the profile's main subject — is why a studio's account posting AI-touched project films is a different case from an account built around an AI spokesperson. TechCrunch also reported the rule doesn't apply to creators who use AI "to edit photos, polish captions, create graphics, or make other creative tweaks." A studio's Instagram account represents the firm, not a fictional person; the small original fictional cast inside an Arch2Video film appears within individual videos, not as the account's identity. Posting a project film that includes people doesn't turn your studio's account into an AI-generated profile under this rule.
The two labels can still overlap on a single post. An individual boosted ad still gets its own AI info label if the underlying video was AI-generated, regardless of your account's profile-label status. The account-level rule and the ad-level label answer different questions, and a studio can satisfy both simply by not misrepresenting who is behind the account.
A pre-boost checklist for this quarter
Before you put part of this quarter's marketing budget behind a rendering or project film, run through this short list:
- Preview the ad after it goes live and check the three-dot menu for AI info, so you know what a viewer sees.
- If the creative includes a photorealistic AI person, plan for it to possibly appear next to "Sponsored," not just in the menu.
- Keep a plain caption noting when a project is a concept study rather than built work, using wording like the examples in our AI rendering disclosure guide.
- Don't rely on cropping or color-correcting alone to avoid the label if the underlying image or video came from an AI tool.
- If the ad is Reels-format, check current Instagram Reels specs alongside this label so the creative satisfies both sets of rules.
If you generate the underlying video with an AI tool, a person should still review every clip before it goes into an ad: generated motion interprets your source images and can change small details, and a fictional cast in the film means people from your original photos won't appear as themselves. That review catches problems a viewer would flag long before Meta's label does. Our own example films are labeled as concept studies for the same reason a boosted ad needs an honest caption.
Frequently asked questions
Will Meta reject my ad for using an AI-generated rendering or video?
Not automatically. The AI info label is a disclosure, not a rejection. Common Thread Collective reported that undisclosed AI content is an active reason ads get rejected, so the risk comes from hiding AI use rather than from using it. A properly labeled AI-touched ad, whether Meta detects it or you disclose it in Ads Manager, can still run normally.
Does the AI info label reduce how many people see my ad?
Meta hasn't published a stated reach penalty for the ad-level AI info label itself. That's different from the separate account-level AI-generated profile rule, which does reduce reach for unlabeled AI-person profiles. Treat the ad label as a transparency requirement rather than a ranking signal, and focus on disclosing accurately instead of trying to dodge the label.
Does a person in my Arch2Video project film make my studio's account an "AI-generated profile"?
No. TechCrunch reported the August 31, 2026 rule applies to accounts whose main subject is an AI-generated person, not to individual videos that include AI-generated figures. A studio account representing your firm, posting project films that happen to include a small fictional cast, is a different case from an account built entirely around one AI persona.
Do I need to disclose AI use if I only resized or color-corrected a photo for an ad?
No. Meta's help center states that minor changes like image resizing or color correction don't trigger the AI info label. The label targets content created or significantly edited with generative AI, such as background generation, image generation, added animation, or third-party AI tools that Meta detects through embedded metadata.
Do third-party tools like an AI photo editor or AI video generator trigger the label too?
Yes. Meta's system reads industry-standard signals, including C2PA Content Credentials that many AI tools embed in exported files, and applies the AI info label when it finds them. That covers AI photo editors and AI video generators used outside Ads Manager, not only Meta's own Background Generation, Image Generation and Add Animation features.