Google is rolling out a new layer of transparency for people who see ads across Search, Discover, and YouTube: a clear label that tells you when an ad was created or edited using AI. The change lands inside Google’s “My Ad Center,” in the same place where users already go to manage ad preferences, block certain advertisers, or report ads. But this time, the focus is narrower and more specific—less about targeting and more about authorship. In other words, Google is trying to answer a question that has become increasingly hard to interpret in the age of generative tools: not just who paid for the ad, but how it was made.
The update adds a “created or edited with AI” disclosure under the “how this ad was made” tab. For users, it’s designed to be discoverable without requiring them to hunt through settings. When you view an ad, you can tap the three dots or the info button to open the panel that explains the ad’s context. That panel is already familiar territory for many people; it’s where you can block or report ads. Now, it also includes the AI label, making the disclosure part of the same quick decision flow—see something, understand what it is, and take action if you want.
This matters because AI-generated advertising isn’t a single phenomenon. It can mean a wide range of practices: generating entirely new creative assets, rewriting copy, swapping backgrounds, producing variations at scale, or editing existing images and videos to make them more attention-grabbing. Some of these changes are subtle enough that a viewer might assume they’re looking at a normal marketing asset. Others are dramatic—hyper-polished visuals, unnatural textures, or language that feels “too perfect” in a way that resembles machine output. Until now, most people had no reliable way to know whether the ad’s creative process involved AI at all. Google’s label doesn’t solve every ambiguity, but it gives users a baseline signal that can be used to interpret what they’re seeing.
What makes the rollout notable is the scope and the placement. Google isn’t only adding a label somewhere in a help page or burying it in policy documentation. It’s integrating the disclosure into the ad experience itself, across multiple surfaces where ads are a daily part of browsing. Google Search is where intent is often explicit—people are actively looking for something. Discover and YouTube are different: they’re more about discovery and recommendation, where the line between content and promotion can feel thinner. By placing the label in all three, Google is effectively saying that AI transparency should travel with the ad wherever it appears, not remain confined to one product.
There’s also a practical reason for this approach. If disclosures are too hard to find, they don’t function as transparency—they function as compliance theater. Google’s design choice suggests it wants the label to be visible at the moment of consumption. The “how this ad was made” tab is a natural fit because it already answers questions users implicitly ask: Why am I seeing this? Who is behind it? Can I do something about it? Adding “created or edited with AI” turns those questions toward the creative pipeline, not just the targeting logic.
Google says it will automatically apply the AI label to ads made with its own generative AI advertising tools. That’s an important distinction. It implies that Google has visibility into the creative workflow when the ad is produced using Google’s internal systems. In that scenario, the label can be applied reliably without requiring manual intervention from advertisers. But the company also acknowledges a reality of the ad ecosystem: not all AI-generated creative originates inside Google. Ads can be produced by third-party agencies, in-house marketing teams, or external vendors using tools that aren’t part of Google’s generative stack. For those cases, Google indicates that the labeling will need to be applied manually.
This split—automatic for Google-made AI, manual for elsewhere—creates a tension that will likely shape how effective the labels are in practice. Automatic labeling is consistent and scalable. Manual labeling depends on advertiser behavior and process discipline. If some advertisers forget, misunderstand, or choose not to apply the label, users may still encounter AI-influenced ads without the disclosure. That doesn’t make the system useless, but it does mean the label is only as complete as the incentives and enforcement around it.
Still, even partial coverage can shift user expectations. Once people learn that an AI label exists, they may start treating it as a meaningful indicator rather than a vague promise. Over time, that can influence how advertisers plan their creative workflows. If the label becomes a standard part of the ad interface, advertisers may begin to treat AI usage as something that needs to be documented and justified—not just something that happens behind the scenes. That could lead to better internal tracking, clearer creative guidelines, and more deliberate decisions about when AI is used and how it’s presented.
There’s another angle here: transparency is not only about informing users—it’s also about shaping trust. In the last few years, generative AI has changed the economics of content creation. It can reduce costs, speed up iteration, and enable personalization at a scale that would have been difficult before. But it also introduces new risks: misinformation, impersonation, and deceptive creative that mimics authenticity. Advertising sits at the intersection of persuasion and credibility. People are already skeptical of ads; they’re trained to expect exaggeration. When AI enters the picture, skepticism can intensify because viewers may worry that the ad is not merely persuasive, but synthetic—constructed in ways that are hard to verify.
Google’s label is a small but direct response to that concern. It doesn’t claim that AI-labeled ads are necessarily misleading, nor does it imply that non-labeled ads are trustworthy. Instead, it offers a piece of context that helps users calibrate their interpretation. A viewer might decide to scrutinize an AI-labeled image more closely, or they might pay attention to whether the ad’s claims align with what they know about the product. Even if the label doesn’t prevent deception, it can reduce the “unknown unknowns” that make deception easier.
The update also reflects a broader industry shift. Regulators and platforms have been grappling with how to disclose AI involvement in content. Some approaches focus on watermarking or provenance systems. Others focus on labeling at the point of display. Google’s method is firmly in the latter category: a user-facing disclosure tied to the ad itself. That’s likely because ads are already heavily regulated in terms of disclosure and user controls. Platforms have long provided mechanisms for reporting, blocking, and understanding ad context. Adding an AI label fits into that existing infrastructure.
For users, the label’s placement inside “My Ad Center” is particularly relevant. “My Ad Center” is not just a passive information page; it’s an interactive hub. Users can block advertisers, adjust preferences, and report ads. By embedding the AI disclosure in that environment, Google is effectively connecting transparency to agency. If someone sees an AI-labeled ad and feels uncomfortable, they can respond immediately—block it, report it, or simply avoid engaging. That’s a meaningful difference from disclosures that exist only as static text somewhere else.
For advertisers, the label introduces a new dimension to campaign performance and brand perception. Creative is already a major driver of ad effectiveness. If AI-labeled ads become associated with certain styles—hyper-realistic imagery, rapid variation, or unusual editing—users may develop heuristics about what those ads tend to look like. That could affect click-through rates, conversion behavior, and even brand sentiment. Some brands may embrace AI as a tool for efficiency and experimentation, while others may prefer to avoid the label by relying on traditional creative pipelines. Either way, the label creates a new variable in the marketing equation.
It’s also worth considering how this interacts with the “how this ad was made” concept itself. That phrase implies a broader explanation of the ad’s production and context. Historically, “how this ad was made” panels have focused on things like why the ad is shown to you, what signals are used, and who is advertising. Adding AI disclosure shifts the panel from purely targeting-related transparency toward production-related transparency. That’s a subtle but significant evolution: it acknowledges that the ad’s origin story includes both distribution mechanics and creative generation.
In practice, the label may also influence how users interpret the ad’s media. For example, if an ad uses an image that looks unusually polished or stylized, the AI label can help explain why. If the ad’s copy reads like it was generated quickly and then refined, the label provides a clue. This can reduce the cognitive load on users who otherwise would have to guess. Guessing is exhausting, and it often leads to either over-trust (“it looks professional, so it must be real”) or over-suspicion (“it looks weird, so it must be fake”). A disclosure can help users land somewhere more grounded: “I know AI was involved, so I’ll evaluate accordingly.”
Google’s statement that it will automatically label ads made with its own generative AI advertising tools suggests that the company is building a bridge between creative generation and compliance. That bridge is essential for scaling transparency. Without automation, labeling would be too burdensome for advertisers and too inconsistent for users. But the manual requirement for third-party AI introduces a gap that the industry will likely address over time. We may see more standardized disclosure practices, stronger contractual requirements between advertisers and agencies, or improved tooling that helps advertisers determine whether their creative qualifies for labeling.
There’s also a question of how “created or edited with AI” will be interpreted. The phrase is broad. It could include minor edits—like using AI to enhance an image—or it could include full generation of assets. It could cover rewriting copy, transforming video frames, or producing variations. The label is meant to be simple for users, but the underlying criteria likely require careful definition. If the criteria are too narrow, many AI-influenced ads might slip through. If the criteria are too broad, advertisers may feel forced to label ads that use AI in ways that are arguably routine