AI isn’t stealing your traffic. It’s stealing your authority

When I was Chief of Staff at CoinDesk, I was in charge of the publication’s approach to AI. One of the earliest debates our internal AI committee had was about whether we should allow AI to index our articles or not.

Most of the people on the committee thought we should block AI crawlers. While the fury of media copyright lawsuits had yet to begin, the issue had gotten some traction, and it was easy to make the case that we shouldn’t give our content away to AI companies to summarize unless we were compensated in some way.

But one person boldly made the case for the other side: He argued that, if AI becomes the new way people find information, shutting ourselves out of AI services would mean our stories—and more broadly, the ongoing narratives around them—would be cut out of the amalgamated answers that the people using AI would read. We would be conceding that ground to competitors to not just get referrals (which, we knew even then, would be few), but to establish consensus. We would no longer be the authority on the things we write about, at least for those who find information through AI portals.

The cost of silence

Little did I know that largely academic debate at the time would become the centerpiece of the AI conversation nearly three years later. Today, information presence in AI summaries—for brands, for public relations, and for the media—is of great interest, and poorly understood. For publishers, the issues of copyright and compensation are ongoing. But regardless of how those conflicts are resolved, AI has become the primary interpreter of their content for a large and growing audience.

The committee didn’t have a name for it back then, but the idea of taking the opposite course of blocking, and actually encouraging AI to index your content, is now called generative engine optimization or GEO (sometimes the first word is substituted for “answer,” or AEO). When I’ve previously written about GEO, it was mostly in the context of why publishers would even want to do it. After all, if AI is taking your content and summarizing it without sending users to your site, what’s the benefit?

There are reasons, but I think it’s more informative to flip the question around: What’s the cost if you don’t? And that is relinquishing your influence on the consensus around the topics in your domain.

The risk isn’t the loss in traffic—that’s lost anyway. Audiences are turning to AI as their information guides no matter what publishers do. What a publisher risks losing is their role as the chief interpreter of events. By reporting facts and validating claims, journalists have historically set the baseline for others to react to. Without those inputs, AI will paint a poor picture of reality.

The thing is, even if a publisher opts out of AI summarization, there will always be someone else who republishes the information who doesn’t (an important foundational concept of copyright law is that, although works are copyrightable, the underlying facts and ideas aren’t). Except now that set of facts is put through their lens, and that will define the “first draft” that machines reuse. Will the answer be inadequate and incomplete? Probably. But as use of AI increases, it’ll be what most interpret as the truth.

That’s why I think framing AI blocking as an existential dilemma kind of misses the point. Blocking AI from indexing your content means blocking yourself from having a say in what a rapidly expanding portion of the world counts as truth. A publisher prioritizing GEO means finding the value in what can’t be captured by traditional metrics like traffic and time on site. Victory in the new battleground of the AI summary will be measured by a different set of criteria: citations in AI answers, influence on narratives, and long-tail impact on trust and authority.

Shaping truth at scale

None of this is to say publishers should just let the AI companies crawl as much as they want and settle for no compensation. If anything, measuring and showing that your content is the source of consensus is hard proof of how valuable the content is. Lawsuits naturally focus on consent, copyright, and compensation, but the rise of GEO reveals what’s really being contested: Who gets to shape meaning at scale.

Demonstrating how specific content influences AI answers is currently a challenge, but that’s about to change. Led by marketers, PR agencies, and brands, there’s a strong push to better understand GEO and how strategies around content, technical factors, and communication can help AI take notice of certain narratives over others. Like SEO, it will always be more art than science, but by this time next year I suspect the field of GEO won’t look nearly so nascent.

On top of that, AI will be an even bigger informational gatekeeper than it is today. Litigation over compensation is important and necessary, but it shouldn’t keep the media from competing to be included in the new crucible where truth is formed. Journalists may no longer control the interfaces where people get information, but they still control the facts. Asserting that role in an AI world doesn’t mean you stop fighting for a better one.