Google Ads AI label: on-ad overlays only for EU, India and New York

Google's AI label setting always adds a disclosure to the "How this ad was made" section of the My Ad Center panel, and it adds a visible overlay on the ad itself only when the campaign targets the European Union, India or New York. If you upload clips made with Sume, you decide per asset whether to mark them as created or edited with AI.

What the setting does, from Google's page

Google's help page says ads that contain assets you designate as AI-generated or edited get a disclosure in the My Ad Center panel, accessible globally, with details on whether the assets were AI created or edited. For campaigns that target the European Union, India or New York, those ads also include visible overlays.

Three notes on the same page matter for a producer. Google may label your assets in some cases, such as where Google itself is legally required to, or when it gets signals from other platforms. Fully-automated Google features can label assets on your behalf. And in any of these cases, labels cannot be overwritten. The page also says that using the setting does not guarantee compliance with specific regulations and tells you to seek legal guidance.

Which Sume outputs count

A clip from POST /v1/videos is generated media: a text-to-video, image-to-video or reference-to-video render. Treat it as AI created. A clip you made on a Timeline from your own footage, with only trims, fades and a soundtrack added, is not generated by Sume, although any AI-made footage inside it still is.

Google's prompt appears during campaign creation as Review assets, with Manage AI label and two choices: label this asset as created or edited with AI, or do not label it. You can also change it later from the asset details page in Asset studio. The choice is yours, so the useful habit is to decide it before upload and write it down.

Log the decision next to the job

Sume returns a job id and the model on submit, and the poll response repeats them. Keep those next to the Google Ads asset name and your label choice. The short script below submits one render and appends the record to a JSONL ledger. It uses only the standard library.

The ledger is yours. It does not talk to Google Ads, and it does not claim anything about compliance. It just means that when someone asks later which assets were flagged, you can answer from a file.

import json, os, urllib.request

body = {"model": "seedance-2", "prompt": "A paper boat crossing a puddle",
        "duration": 5, "resolution": "720p", "aspect_ratio": "9:16"}
req = urllib.request.Request(
    "https://api.sume.com/v1/videos",
    data=json.dumps(body).encode(),
    headers={"Authorization": "Bearer " + os.environ["SUME_API_KEY"],
             "Content-Type": "application/json"},
    method="POST",
)
with urllib.request.urlopen(req) as res:
    job = json.load(res)
row = {"job_id": job["id"], "model": job["model"],
       "google_ads_ai_label": "label this asset as created or edited with AI"}
with open("ai-asset-ledger.jsonl", "a") as f:
    f.write(json.dumps(row) + "\n")
print(row)

Why location is the part to check

Because the overlay depends on where the campaign targets, adding a New York or EU ad group later changes what viewers see on assets you already uploaded. Review the AI label column in the asset library when you add a location, using the asset filter for AI label.

If a label already exists because Google applied it, you cannot undo it. So decide before the first upload, not after.

What this does not decide

Whether a given law applies to your ad is a legal question. Google's page names regions, not statutes, and Sume does not interpret them. Sume gives you a clean record of which assets came from a generation job.