At the 2026 RampUp conference AI was clearly the dominant topic across almost every session and conversation.
But what struck me most wasn’t the technology itself — it was how early the industry still is in figuring out how AI will actually be implemented.
After two days at the conference, here are three takeaways that stood out to me.
The biggest thing missing from AI in advertising right now is standards.
Scott Howe, the CEO of LiveRamp, gave an interesting keynote where he talked about waves of innovation over history. He mentioned the rise of the railroad in the United States in the 1800s, and how the network effects really started to accelerate once standards were introduced — things like standardized time zones and the gauge of the track.
But as I sat through the sessions at RampUp, I was struck by how early we still are when it comes to standards around AI.
The protocols are either still being developed or are only at the proposal stage. In some cases there are even competing standards emerging — for example AdCP, which is being promoted by Yahoo, Scope3, and PubMatic, and AAMP, which is being promoted by LiveRamp and the IAB.
And to state the obvious, Americans don’t really like standards. We tend to adopt them reluctantly, and often the first version ends up being different from the rest of the world.
So my takeaway is that it won’t be the speed of technology that slows the rise of AI agents in buying and selling advertising.
It will be the speed of development of the protocols around AI.
AI in AdTech currently doesn’t have an organized resistance.
If I compare this wave of innovation to others I’ve seen in AdTech, there has almost always been one group pushing for the change and another group pushing back.
A good example is the rise of programmatic buying. There was a camp pushing for it because it promised efficiency. But there was also a very vocal group — mostly publishers and agencies — arguing that it would lead to lower prices and a worse experience for advertisers.
And of course, many people were also worried about what it meant for their jobs and revenue.
But I’m not seeing that same dynamic yet with AI.
Maybe people have learned from the programmatic transition that you can’t really hold back the tide.
My takeaway is that AI may be the first major innovation wave in advertising where there is almost no organized resistance.
The conversation about AI is much further ahead than the actual implementation.
Yes, almost every post in the industry is about AI.
Yes, almost every conference session is about AI.
But there is still a large gap between talk and action.
In one session the moderator asked how many companies in the room had actually run campaigns through an agentic workflow. Only about 20–25% of the hands went up.
And based on conversations I had in the hallways, I suspect the real number may actually be lower.
Even when moderators asked panelists for specific examples of how they were using AI, the examples were still relatively limited.
My takeaway is that the industry conversation around AI is much further ahead than the operational reality.
One closing thought.
Most of the sessions I attended were focused on AI, but I also sat in one session focused on privacy and the patchwork of laws being created by different states in the U.S.
It is a patchwork — but it is also becoming increasingly impactful.
AI is accelerating execution at the exact same moment that privacy laws are increasing constraints.
Which means the companies that succeed in the future will likely be the ones that are best at governance, control, and structured data.