Google Was Ruled an Illegal Monopoly Twice. You Still Trust Its AI With Your Money.

Here's a sentence that should stop every media buyer mid-scroll: in the space of eight months, two different U.S. federal judges ruled, in two separate trials, that Google was illegally running a monopoly. Not "aggressive." Not "dominant." Illegal. And in that same stretch of time, the entire paid media industry — Korf Digital included — kept doing exactly what it did the day before each ruling: logging into Google Ads, handing an algorithm a budget, and trusting it to spend that money well.

That's not hypocrisy. It's the actual bind the industry is in right now, and almost nobody is saying it out loud. So let's say it.

Two Monopoly Rulings in Eight Months

In August 2024, Judge Amit Mehta ruled that Google illegally maintains a monopoly in general search and search advertising — built and defended largely through the default-placement deals that made Google the water everyone swims in. In April 2025, a second judge, Leonie Brinkema, ruled in a completely separate case that Google illegally monopolized the ad tech stack itself: the publisher ad server, the ad exchange, and the illegal tying between them that let Google sit on both sides of nearly every open-web auction.

The remedies are still being argued. The Department of Justice has floated open-sourcing the auction logic that decides which ad wins, with divestiture of the ad server held in reserve if lighter fixes don't work. Google is appealing both rulings. None of that changes the finding two separate courts already reached: for years, the referee was also a player, and it structured the game to win.

The Black Box You're Not Allowed to Open

Here's the part that should actually keep you up at night: none of this changes how you run a campaign tomorrow morning. Performance Max, AI Max, Advantage+, Andromeda, GEM — the entire industry has spent the last two years quietly converting "how ads get bought" from a set of decisions you make into a set of decisions you request. You submit a budget, a goal, and some creative. Something you cannot audit decides who sees it, when, and for how much. You get a dashboard number back. You are simply not given the option to check its work.

That wouldn't be a crisis if the dashboard number were trustworthy. It increasingly isn't. 75% of marketers now say their measurement systems are broken — not underperforming, broken. When AI-driven bidding decisions can't be explained or traced, confidence collapses at exactly the level that controls budget: CFOs don't care how sophisticated the model is, they care whether they can trust what it's telling them. The IAB was concerned enough to launch an industry-wide initiative — Project Eidos — purely to try to rebuild trust in measurement itself.

The Jobs Math Nobody's Advertising Out Loud

Martin Sorrell — who built WPP into the largest ad holding company on Earth before founding S4 Capital — put a number on this that most of the industry is quietly nodding along to and publicly ignoring: roughly 250,000 people currently work in media planning and buying, and in his words, "there won't be 250,000 jobs in two to three years." His argument isn't speculative — it's already visible in the product roadmaps. Performance Max and Advantage+ made targeting and bidding algorithmic years ago. In February 2026, Meta went further, embedding an autonomous agent directly into Ads Manager to handle report building, audience research, and campaign analysis — the parts of the job that used to require a person asking questions.

Separately, agencies are projecting roughly 15% of their own jobs disappearing by the end of 2026 as automation eats the billable-hour, run-the-dashboard work that used to justify a headcount. The honest read isn't "media buyers are extinct" — it's that the job now consists almost entirely of the parts a black box can't do: deciding what's actually true, questioning a number that looks too good, and building measurement the platform has no incentive to build for you.

So What Do You Actually Do With This

1. Run independent incrementality tests, not platform-reported ones. A holdout group is the only honest way to know what your spend actually caused.
2. Refuse to let one platform grade its own homework. Cross-check Meta's reported conversions against your CRM, not against Meta's own pixel.
3. Diversify the auction you're exposed to. A monopolized ad-tech stack prices differently when it knows you have no alternative.
4. Treat "the algorithm is learning" as a claim, not a guarantee. Ask what it's optimizing for, and whether that's actually your business goal.
5. Budget for measurement like it's media, not overhead. The accounts that will survive this decade are the ones that funded the audit, not just the ad.

The Uncomfortable Part

None of this means walk away from Google or Meta — there's nowhere else with that reach, and both platforms know it. That's precisely the leverage a court just ruled was obtained illegally. The uncomfortable truth isn't that the platforms are evil; it's that an industry built on trust just had two federal judges confirm, in writing, that the trust was misplaced — and the response from most advertisers has been to keep clicking "apply" anyway, because the alternative is learning to do the job the black box was hired to make disappear.

The agencies and in-house teams who make it through this decade won't be the ones who automated fastest. They'll be the ones who kept a human in the room whose entire job is refusing to take the dashboard's word for it.