A $500 billion reminder of how the duopoly wins the internet

A remarkable milestone just slipped by without a peep from industry pundits. In the last quarter, Google and Meta surpassed an annual run rate of one half of a trillion dollars in advertising revenues.

That number should have stopped us in our tracks. It did me.

$500 billion: Literally, half of the global advertising market.

It has been over a decade since DCN first identified the "duopoly" as a force which, if left unchecked, would increasingly extract value from publishers while capturing over 90% of the industry growth quarter after quarter after quarter. There have since been regulatory actions, congressional hearings, privacy scandals, whistleblowers, and countless debates about the power of the dominant platforms. Entire industries, including media winners and losers, have been ransacked and reshaped based on their decisions. Yet, it's hard to argue that this course of dominance and destruction has changed in any meaningful way.

The courts are still working through cases that began years ago. The FTC continues its efforts against Meta. Google remains entangled in multiple antitrust actions. Meaningful remedies, if they come at all, will likely take years more.

But the money continues to flow.

Antitrust efforts fail to rectify internet economics

That half-trillion-dollar milestone tells us something important. Despite a decade of scrutiny, remarkably little has changed about who captures the value of media created across the open internet.

One of the more surreal moments for me came in late 2018 when Google contacted me personally to explain why Facebook was different and the real problem. The irony wasn't lost on me then, and it isn't now. Though the two companies publicly positioned themselves as rivals, we later learned that they were allegedly secretly working together to rig the ad market and damage publishers. CEO Sundar Pichai's 2025 deposition regarding his discussions with Mark Zuckerberg remain under seal. If not for the Texas attorney general's ongoing case, we might never see it. Time will tell.

Yet the industry press has largely moved on. Last week's FTC appeal in its effort to unwind Meta's acquisitions of Instagram and WhatsApp barely registered. Perhaps that's understandable. Court proceedings move slowly, and the media industry does not have that luxury.

How audience intelligence became the internet's most valuable asset

But none of this has really been about advertising.

Advertising was simply the funding mechanism.

The real issue has been about who creates value and who captures it. As we wrote back in 2016, the dominant platforms built their market power through unprecedented data collection. They track users across their own properties, but more importantly across the broader web. As content is consumed, they harvest signals continuously - learning where we go, what we read, what we watch, what we search for, and what we might do next to fuel advertising and search results.

Publishers made the investments, but it was the platforms that captured these signals.

That distinction matters.

News orgs invest in reporting and journalists. Entertainment companies invest in storytelling and creators. Together, they invest in the content, experiences, and environments that attract audiences and earn trust.

The platforms just discovered something even more profitable than creating content: what they could observe about the people consuming it.

Every search, click, view, purchase signal, mobile location, and behavioral pattern became part of an expanding system of surveillance and targeting. The value being captured was not the content itself. It was the relationship between audiences and content. It was the context surrounding attention; the intelligence generated whenever consumers engaged with trusted brands. The more content was created across the open web, the more opportunities to collect behavioral information to tune advertising products and monopolized distribution feeds.

One side created value while the other side built extraordinarily profitable businesses around efficiently observing and monetizing that value.

Back in 2016, DCN was the first org willing to challenge this model directly. We argued that surveillance-based business models were creating structural distortions weakening the economic foundations that support trusted news and premium content. At the time, many viewed those concerns as alarmist. Others dismissed them as anti-technology.

A decade later, many of those same concerns now sit at the center of antitrust cases around the world. Yet as the legal system catches up to the first era of the internet, the industry is already confronting the next.

AI is being built on the broken economics of the platform era

Speaking this week at the World News Media Congress, New York Times Publisher AG Sulzberger warned that AI companies are misappropriating copyrighted works at an unprecedented scale while threatening the economic framework that supports the news media business. While his focus was on AI, the pattern he described was very familiar. In many ways, it represents the next chapter of the same story.

Yes, without a doubt, the next phase is AI. And it is unleashing amazing opportunities that publishers are leaning into. But it's also attempting to execute on the same playbook of ransacking "data" with impunity and monetizing it somewhere else more profitable to the platform. If the first phase of platforms harvested the value created in front of and around the content, then this AI phase increasingly seeks to harvest the content itself.

And many of the angels born of the open web are now pivoting towards profiting from AI, too.

Which brings me to Common Crawl.

Today, DCN sent a formal legal notice to Common Crawl, one of the largest repositories of web content used throughout the AI ecosystem. The notice raises serious concerns regarding the collection, retention, and redistribution of publishers' protected content without authorization. It challenges a growing assumption that content created through substantial investment can be collected, stored, repurposed, and monetized simply because it is technically accessible. And just last week Common Crawl announced a deal with Hugging Face to make that embrace of enabling copyright infringement by for-profit AI companies and developers even more seamless.

However, this is not really about a single organization. It is about a mindset that has become embedded throughout parts of the AI ecosystem.

If it can be scraped, it can be used. If it can be used, it can be monetized. The dominant platforms have spent years reducing users, context, and content into a single catch-all term: data. Once everything becomes data, the creator, the brand disappears from the equation. Or maybe more accurately, it can be hijacked into an algorithm at scale without the undesirable expense of journalists, storytellers, institutional trust and liability.

Remember: We'll see $500 billion extracted from this "data" this year alone.

Journalism is not data. Articles are not data. A video is not data. These are investments in expertise, human creativity, and trustworthiness. They are assets created through risk, talent, and capital. Data is merely the binary bits of the internet underneath them. By reducing everything to data, the act of content creation vanishes, and the question of who deserves to be compensated becomes easier to overlook.

That is why the debate surrounding AI shouldn't be merely a technology debate. It is an economic debate. It's a debate about whether those who create value will continue to get the short end of the stick to those who capture it.

Sulzberger's warning is therefore about much more than journalism. It is about whether society still recognizes where value originates. As he noted in his remarks, less than 1% of the AI investment flows to the content that trains these systems, while hundreds of billions are being directed toward the three other basic ingredients: energy, compute, and engineering talent.

Answering the unresolved question at the center of the digital economy

For 25 years, the digital economy has often rewarded those who capture value much more richly than those who create it. The antitrust battles involving Google and Meta may eventually address some of the excesses of the first internet era. The copyright, licensing, and governance battles surrounding AI will help shape the second.

The first era monetized our behavior. The next aims to replace our content.

Both stem from the same unresolved question:

Will we continue treating human creativity, expertise, and audience relationships as raw material to be harvested by others? Or will we finally recognize that every advertising system, every dataset, and every AI model ultimately depends on people and trusted institutions willing to create original value in the first place?

The half-trillion-dollar advertising milestone should not be viewed as a business achievement alone. It is a scoreboard. Yes, the emerging technology boom may be different. But until we answer that question, little else will change.