Inside ChatGPT’s Ad Test: Intent-Led Discovery, Limited Scale, and the Cost of Early Access

ChatGPT’s long-anticipated move into advertising is finally underway. After months of speculation, OpenAI quietly began testing ads inside ChatGPT in early February. The rollout has been deliberately slow, with limited placements, minimal spend, and a small group of participating brands. For an industry used to fast-moving launches and aggressive scaling, the pace has been noteworthy.

Beneath the slow start is something more interesting: a new model for how advertising might work inside AI-driven environments. Early signals suggest this won’t look like traditional search or social. Instead, it’s shaping up as a more controlled, context-driven experience, and one that could change how brands show up when consumers are actively asking for help. Here are three early takeaways for advertisers:

• Right place, right time: The biggest difference with ChatGPT ads is where they appear. Instead of interrupting scrolling or competing for attention in a feed, ads are tied directly to user intent. Someone comparing smartphones might see a Best Buy placement while someone planning a trip might get an Expedia recommendation. This shift changes the role advertising plays. Rather than trying to capture attention, brands are showing up inside moments where users are already asking questions and making decisions. It’s closer to assisted discovery than traditional media, which opens the door for ads that feel more relevant, but it also raises the bar for usefulness. If the placement doesn’t add value, it’s much easier for users to ignore or distrust it.

• Trust sets the pace: So far, OpenAI has kept a tight grip on the rollout. Spend has been limited, reporting is basic, and many advertisers are still waiting for broader access to tools like self-serve buying and real-time performance data. While that’s created some frustration, it also reflects the stakes. ChatGPT operates in a more personal, trust-based environment than most ad platforms. Users aren’t just browsing, they’re asking for advice, planning purchases, and making decisions. Introducing ads into that experience requires a different level of caution. The result is a slower, more controlled rollout that prioritizes user experience over immediate revenue. For advertisers, it’s a reminder that this won’t scale like traditional platforms overnight.

• Early access trade-offs: More than 100 brands have already tested ads in ChatGPT, drawn in part by the opportunity to establish an early presence. Some are already seeing promising signals, including increased traffic and engagement tied to AI-driven discovery, but the experience isn’t fully built out yet. Reporting is still evolving, customization is limited, and the buying process lacks the sophistication marketers are used to. In many ways, it's still very much in a pilot phase. This creates a familiar dynamic: early movers get access and learning, but with fewer tools and less control. The value isn’t just in immediate performance, but in understanding how this new environment works before it scales.

If this model sticks, advertising will start to take on even more of a participatory role. Instead of inserting messages into a feed, brands are entering moments where users are actively asking for help and expecting useful answers. In this new environment, brands aren’t just competing for attention, they’re competing to be part of the answer. That’s a very different game to play, the rules of which are just beginning to be written.