Fifteen years ago China Mobile published a white paper about saving electricity. That is really where this started, with a document arguing you could pull the baseband out of the towers, pool it in a central room, and cut the power bill. I wrote about that C-RAN paper as the origin of everything that came after, and standing here in the middle of 2026 I keep thinking about how far the idea travelled from that modest, almost boring starting point.

From there it became a software bet, and then a body writing specifications, and then a greenfield network in Japan that proved the thing could actually run. After that it turned into geopolitics, spread across Europe as a diversification campaign, and reached the incumbents when AT&T handed Ericsson fourteen billion dollars and called it open. Every one of those was a real step. Add them all up and you land on something like ten billion dollars of cumulative Open RAN spending and a movement that is, by any fair reading, mature.

And here is the number that has bothered me through this entire series, the one I kept circling without quite landing on it. After fifteen years, the RAN market is more concentrated than it was before the O-RAN Alliance existed. Supplier concentration is up in nearly every region on the planet. The single thing Open RAN was supposed to deliver, more suppliers and less dependence on any one of them, is the thing it most clearly did not. The AT&T deal was the tell in miniature, an open contract whose first effect was to push a competitor out and consolidate onto a single vendor. Multiply that instinct across a whole industry and you understand how the market got more concentrated while everyone was talking about opening it up.

I do not think that makes it a failure, though, and I have gone back and forth on this more than once. The open interfaces are real now, sitting in the specifications and in shipping equipment, and the actual innovation stopped being about radios and moved up into the software, into the RIC and the xApps and the AI-driven optimization everyone is demonstrating at every trade show. More than that, the whole architecture is quietly being folded into 6G instead of bolted onto 5G as a retrofit. John Strand, who has been the loudest skeptic for years and whose adoption numbers were mostly right, has a line that keeps looking sharper: Open RAN may end up being a 6G architecture rather than a 5G one. That is a strange kind of win. It did not diversify 5G. It might simply be how 6G gets built.

You can already see it if you look at the right thing. I wrote last week about Viettel and Qualcomm starting serious 6G work, and the detail worth noticing is that nobody bothered to frame it as an Open RAN announcement. The open interfaces were just assumed, part of the furniture, not worth a bullet point. That is what winning quietly looks like, a set of ideas so absorbed that people stop naming them.

So I am left somewhere I did not expect. Openness became the default vocabulary of the industry and, at the same time, failed at the specific promise it was founded on, and I cannot decide whether that is the movement succeeding or the industry swallowing it whole and neutralizing the part that threatened anyone. Fifteen years, from a white paper about electricity bills to a ten-billion-dollar line item and an assumption baked into the next generation, and the one metric that was meant to prove the whole thing worked, how many vendors a carrier can actually choose between, moved the wrong way the entire time. I still do not have a clean answer. I am not sure the industry has one either, and I will keep watching to see which reading turns out to be true.

Sources