Google's real moat was never the algorithm. It was the data. Two decades of watching what billions of people searched, clicked, and skipped taught its systems things no competitor could reproduce, because no competitor had the traffic to learn from.
That moat just got a legal crack in it. Under the Digital Markets Act, the European Commission has adopted binding decisions requiring Google to share anonymized Search data with rival engines and AI chatbots that qualify as eligible providers.
A recent piece on Search Engine Journal laid out the mechanics. Read past the legal language and the strategic picture is stark. The thing that made Google uncatchable is being handed, in part, to the people trying to catch it.
What Actually Changes Hands
The data in scope is specific. Google must provide anonymized query, click, view, and results-position data. It does not have to hand over account details, individual search histories, or the ranking algorithms themselves.
So this is not Google's secret sauce being copied wholesale. It is the raw behavioral signal, what people ask and what they click, that any search or answer engine needs to learn what a good result looks like. That signal is the hard part to acquire from a standing start.
The access is gated, not open. A qualifying provider needs at least 50,000 monthly EU users, either a two-year operating history or a passed investment test, plus security screening and independent audits. This is aimed at serious challengers, not weekend projects.
The timeline is concrete too. Google must build the dataset and set pricing by January 2027, with Android AI changes required by August 1, 2027. This is not a distant white paper. It is a countdown with dates on it.
The obvious beneficiaries are AI answer engines. Being able to ground responses in real search behavior, and to earn referral traffic from it, is exactly what the current crop of chatbots lacks. This lowers that barrier for anyone who clears the bar.
Look at what the exclusions protect. Google keeps its ranking algorithms and its individual user histories. So the settlement hands over the fuel, the aggregate behavior, while keeping the engine, the secret method, locked. That is a careful line, and it tells you exactly what Google believes its real edge is.
Why This Matters Even If You Never Touch the Data
Most readers will never file to access a byte of this. The point is what it signals about where discovery is heading.
For twenty years, one company decided what visibility meant, because one company owned the surface where discovery happened. If you wanted to be found, you optimized for Google, full stop. That single-surface reality is ending, and regulation just accelerated it.
When more engines and chatbots can learn from real behavioral data, discovery fragments across more surfaces. Your customers stop starting every search in one box. They ask an AI assistant, a vertical engine, a chatbot inside another app, and each one is now better equipped to give a grounded answer.
I wrote in Google keeps two thirds of its clicks now about how the giant hoards attention. This is the counter-move at the structural level. The state is trying to redistribute the input that made the hoarding possible.
The Strategic Read for Brands
Stop building your visibility strategy as if Google is the only weather system that matters. It has been the safe assumption for a generation. It is becoming a risky one.
The grounding layer is where this gets practical. When multiple engines can pull from strong behavioral and source data, the brands that win are the ones whose information is clean, consistent, and easy for a machine to trust across every surface, not just Google's. I made that case in Microsoft just decided which sources AI trusts, and this ruling widens the field of engines doing that trusting.
Treat discovery as a portfolio, not a single channel. Measure where you appear across the engines your buyers actually use, and assume that list is going to grow, not shrink. The monopoly on the front door is being unbundled by law.
There is a caution worth holding too. The article notes that AI platforms still represent a small slice of global traffic today, so the immediate effect of this ruling is limited. Do not overreact and abandon a channel that still sends most of your visitors.
The point is direction, not speed. Regulators rarely move to unbundle an asset unless they already see it as the chokepoint of the next decade. The EU is not fighting the last war here. It is trying to shape who gets to build the answer engines that come after search, before one company owns that layer the way it owned the last one.
None of this means Google is finished. It still owns the habit, the defaults, and the scale. But the argument that nobody could ever compete because nobody could get the data no longer holds. The moat is being drained on a public schedule, and smart brands should plan for the day the water is gone.