AI Ad Labels on Meta and Google: India Agency Checklist (2026)

Direct answer

Indian agencies should treat AI-ad labels as a production-control and evidence-capture issue, not a verdict on creative quality. Google now provides advertiser label controls and “How this ad was made” disclosures, while Meta is rolling out “About this ad” with AI information. Record the label, region, asset provenance and approval owner. Presence or absence of a label proves neither legality nor performance.

What changed in Google Ads?

Google announced additional AI-ad transparency features on July 9, 2026. Its “How this ad was made” section can appear in My Ad Center when an ad uses AI-created or AI-edited assets. Google says the disclosure is accessible through the three-dot menu on ads across Search, YouTube and Discover.

Google’s advertising products can apply a disclosure automatically to certain assets generated with Google tools. Advertisers can also designate externally created assets as AI-generated or edited. The AI label setting is rolling out across Google Ads, Display & Video 360, Campaign Manager 360, Merchant Center and Google Ads Editor.

For campaigns targeting India, the European Union or New York, Google says designated AI-created or edited assets can also receive a visible overlay on the ad. Google explicitly warns that using its AI label setting does not, by itself, guarantee compliance with a specific regulation. Review the current Google Ads label instructions and My Ad Center explanation before documenting a client workflow.

What changed in Meta ads?

Meta’s June 1, 2026 update says it is beginning to roll out “About this ad,” a unified destination in the three-dot menu. Meta says this surface can include “AI info” labels already applied to ads created or significantly edited with Meta’s generative AI tools.

Meta also says it will use industry-standard signals to detect ads created or edited with third-party AI tools. When detected, it can add an AI information label. Meta notes that the experience may vary in some regions because of legal requirements.

The placement of a label depends on the nature of the edit. Meta says a significant edit made with its tools can produce a label in the menu or next to “Sponsored.” If the result includes an AI-generated photorealistic person, the label appears next to “Sponsored.” Read Meta’s current generative AI ads transparency update.

Why should Indian agencies keep a separate disclosure workflow?

India’s February 2026 amendments to the Information Technology Rules added due-diligence provisions for synthetically generated information. The provisions address intermediaries and computer resources that enable creation, modification, publication or sharing of synthetic content. They include prominent labelling and permanent metadata or other provenance mechanisms where technically feasible. MeitY’s FAQ says the defined synthetic-content category primarily covers audio, visual and audio-visual material that is made to appear authentic; pure text alone is outside that specific definition.

This creates two distinct agency questions:

1. What disclosure does the advertising platform provide or require?
2. What records and legal review does the client need for the asset and market?

Do not collapse these questions into one checkbox. A platform control is evidence of a platform action. It is not a legal opinion. This article is an operational research guide, not legal advice. Agencies should obtain qualified advice for a client’s facts, sector and campaign geography.

Original artifact: AI disclosure evidence and control card

Use one card per material asset, not one card for an entire account.

The public and private layers should remain separate. Public competitor research records what an outside observer can see. The private control layer records facts available only to the client and its agency.

How should competitor-ad researchers record AI labels?

Treat the label as one observable field in the ad creative analysis rubric. Capture it alongside the offer, proof, format, advertiser identity, source and observation date.

For cross-platform comparisons, use the Meta and Google ad-intelligence framework. Do not code Meta’s “AI info” and Google’s “How this ad was made” as identical events. The platforms describe different triggers, placements and detection methods.

When turning a labelled competitor ad into a hypothesis, keep the label out of the performance argument. The research-to-brief workflow should still require an original angle, brand-owned proof and a first-party test.

Advertisng’s research methodology provides the wider evidence standard: record what is public, state what is inferred, and preserve what cannot be known.

What can an AI-ad label prove?

A visible label can prove that a particular disclosure appeared in the observed placement at the recorded time. Depending on the platform, it may also show that the advertiser designated an asset, the platform generated or detected it, or another signal triggered the disclosure.

It does not automatically prove:

• which model or prompt created the asset;
• how much of the asset was generated or edited;
• whether every AI-assisted element was detected;
• whether the advertiser owns the necessary rights;
• whether the disclosure satisfies every applicable law;
• whether the creative is truthful, original or safe; or
• whether the ad produced conversions, revenue or profit.

The absence of a visible label does not prove that AI was not used. Google says not all AI-created content requires a label, and Meta’s treatment depends on factors such as the type and significance of the edit. Regional and product rollouts can also differ.

What can public ad data not prove?

Public ad data cannot prove a competitor’s prompt, generation history, consent records, rights clearance, targeting, budget, bid strategy, conversion rate, revenue, profitability or causal performance. A label is provenance context under a platform’s current system. It is not a performance badge or a complete compliance record.

Frequently asked questions

Does every AI-assisted Google ad need a label?

No. Google says requirements vary by region and type of content, and not all AI-created content needs to be labelled. Agencies should check the current campaign, asset and market guidance rather than apply one universal rule.

Does Meta label ads made with third-party AI tools?

Meta says it is beginning to detect third-party AI use through industry-standard signals and can include an “AI info” label in “About this ad.” Detection and visibility should not be assumed for every asset or region.

Does a platform label guarantee compliance in India?

No. Google explicitly says use of its AI label setting does not guarantee compliance with a specific regulation. Agencies should preserve production records and obtain legal advice where required.

Can a missing label be recorded as “human-made”?

No. Record it as “no visible label observed,” with the platform, placement, region and date. That is the evidence available.

Can public AI-label data identify high-performing competitor creative?

No. Labels do not expose delivery, conversions, profit or incrementality. They should not be used to rank ads as winners.

A restrained next step

Use the control card to manage your own AI-assisted assets. If the agency also needs a repeatable place to save visible Meta and Google competitor evidence, the Advertisng Ad Library can support that research layer. Keep private production records and public competitor observations separate.

Primary sources

1. Google Ads & Commerce Blog: Expanding AI transparency in ads, July 9, 2026.
2. Google Ads Help: Use AI content label settings and disclosures, reviewed August 12, 2026.
3. My Ad Center Help: AI Transparency in ads, reviewed August 12, 2026.
4. Meta Newsroom: Expanding GenAI Transparency for Meta’s Ads Products, updated June 1, 2026.
5. Ministry of Electronics and Information Technology: Information Technology Rules updated February 10, 2026.
6. Ministry of Electronics and Information Technology: FAQ on the 2026 synthetically generated information amendments.

Cover-image brief

Create a 1600×900 editorial illustration using Advertisng’s dark charcoal palette with restrained green, amber and blue accents. Show two public-ad evidence panels labelled “Meta” and “Google” feeding into an India agency approval card with three fields: “Label,” “Provenance” and “Owner.” Use the short headline “AI Ad Labels” and small label “India Agency Checklist.” Avoid platform logos, legal claims, performance claims, biometric imagery and imitation product interfaces.

Suggested alt text: “India agency checklist linking Meta and Google AI-ad labels with provenance and approval records.”