AI in Advertising: What Google and TikTok Ask of AI-Made Ads

Google and TikTok both accept ads made with AI, and each adds a disclosure step. Google posted on July 9, 2026 that it permits labels inside image and video creatives that were generated or modified using AI Source [1]. TikTok says an ad with undisclosed AI content that it finds will be rejected or restricted Source [3].

This page reads two platforms' help pages as they stood when we fetched them. It does not cover Meta's ad rules, and our guide to the Instagram AI label covers how Instagram labels posts. It is general information, not legal advice.

These rules are for paid ads, which Autonomah does not buy: it works without paid ads and builds the brand's communication on its own networks Source [5].

What Google says

The July update says the labels will not violate Google's policies on text overlays and watermarks Source [1]. Google Ads may also apply labels on its own to some assets made with Google AI tools Source [1]. An advertiser can add a label to the creative or use an AI label setting, which Google says will launch gradually throughout July in Google Ads, Display & Video 360, Campaign Manager 360, Merchant Center and Ads Editor Source [1].

The reason it gives is regulation. The page says rules in the European Union, India and New York require ads with certain AI-generated or edited assets to carry disclosures or labels Source [1]. It also warns that using the label setting does not guarantee compliance with a specific regulation, and it tells advertisers to seek legal guidance Source [1].

Election ads have a separate rule. Advertisers must disclose every election ad that contains synthetic or digitally altered content by selecting the "Altered or synthetic content" checkbox in their campaign settings Source [2]. The policy leaves out edits that do not matter to the claims in the ad: resizing, cropping, color or brightness corrections, defect correction such as red-eye removal, and background edits that do not create realistic depictions of actual events Source [2]. The July update tells election advertisers to keep using that checkbox Source [1].

What TikTok says

TikTok's policy on misleading and false content allows significantly edited media and AI-generated content when the advertiser applies the AIGC label, or adds a clear disclaimer, caption, watermark or sticker of their own Source [3].

The policy calls content "significantly modified by AI" when AI changes it beyond minor tweaks. Its examples are images, video or audio that are completely AI-generated, a primary subject shown doing something they did not do, and a primary subject made to say something they did not say with AI voice-cloning Source [3]. Adjusting lighting, brightness or color saturation, removing or changing a background and denoising count as insignificant edits Source [3]. When you are unsure, the page says to use caution and label the content Source [3].

Inside Ads Manager, the help page for AI-generated content disclaimers tells you to check the toggle "This ad contains AI-generated content" while creating the ad Source [4]. The feature covers TikTok ads and not Spark ads, which follow the Community Guidelines instead Source [4]. After you submit, the toggle cannot be changed unless you change the video or image, and duplicating a campaign resets it Source [4].

The same policy does not allow content that misuses a public figure's likeness, audio or visual, without permission, particularly in misleading contexts Source [3].

A label is one requirement among several

TikTok's page keeps its other rules in force. Ads must not promise or exaggerate results, and "Get slim legs right away." is among the examples it lists as not allowed Source [3]. Nothing on the page says a label offsets those rules. Our read: treat the label as a step you add, not as a pass for the claim.

A routine for AI-assisted creatives

This is our own routine. Neither platform prescribes it.

1. Keep one line per creative: what tool made or changed it, and whether it touches image, video or audio.
2. Mark any creative that is completely AI-generated or that shows a person doing or saying something they did not, since those match TikTok's examples Source [3].
3. Set the TikTok toggle while you create the ad, because it cannot be unselected later unless you change the video or image Source [4].
4. Re-read both help pages before each campaign. Google's page is dated July 2026 and TikTok's April 2026, and both can change Source [1]Source [3].

Whatever Autonomah makes belongs to the brand and carries its identity Source [5]; our advice, which the site does not give, is to keep a note of which tools touched each piece next to its file.

How Autonomah does it

Autonomah is an artificial creative and advertising brain that thinks, creates and publishes each brand's advertising Source [5]. In the site's words, a content tool does what it is asked, and Autonomah starts with the brand's diagnosis and proposes what to say, to whom and why Source [5]. You can talk about your brand with Autonomah.

Frequently asked questions

Do I have to label AI-generated ads?

It depends on the platform and the ad. TikTok allows significantly edited media and AI-generated content with a label or your own clear disclaimer, and rejects or restricts undisclosed AI content it finds Source [3]. Google requires the "Altered or synthetic content" checkbox for election ads with synthetic content, and permits labels inside other creatives Source [1]Source [2].

Does Google label AI ads for me?

Sometimes. Google Ads may apply labels automatically to certain assets generated by Google AI tools Source [1]. The page does not say that this covers every AI-made creative, and it does not say that Google applies a label to anything an advertiser made elsewhere Source [1].

What counts as a minor AI edit on TikTok?

TikTok's examples of insignificant edits are adjustments to lighting, brightness or color saturation, removing or modifying backgrounds, and denoising Source [3].