I do not believe Intel was given a chance to stop it.
Intel had their shot to try and convince Dr. Cutress that the NDA represented a lasting and meaningful relationship between Intel and Anandtech, and that basically if Anandtech valued review samples from any hardware company (not just Intel) that end-running the NDA by buying a commercially available chip "accidentally" sold pre-NDA from a retail (read: non-QS/ES) stockpile represented a departure from what a "big review site" is meant to do in this day and age. Intel brings a lot of technical support to the table to respected reviewers to try to help them get an optimal review platform up and running for launch-day articles. By just buying a random retail sample, Anandtech forgoes all of that and is stuck reviewing the hardware on its own, which is very late 90s/early aughts. And it's kind of refreshing. Still there's at least the possibility that Intel will be mum on helping AT get their NDA-lift review sorted wrt UEFI/software updates.
You have to think that if Dr. Cutress actually thought Intel would be wounded by his decision to review a leaked retail sample, that he would probably sit on his (p)review of the 11700k. Or he would launch it at the same time as his review of the NDA-addled sample from Intel to do a compare-and-contrast to show what, if anything, Intel had been able to change over the course of a month by way of UEFI updates. Fact is, Intel said . . . nothing. At all.
Meanwhile, sites like Anandtech are (or were) getting scooped by randos on Youtube publishing results from their "accidental" 11700k retail samples. Everyone on the NDA train had to sit idly by and watch themselves look irrelevant in comparison. Who's going to line up for NDA-lift articles when reams of data on these CPUs is already available to the public? Obviously, the touches we expect from seasoned hardware reviewers were missing in favor of just some benchmark scores produced with configurations that may or may not inform us as to exactly how much better (or worse) the 11700k is compared to the 5800x or 10700k . A good review site like Anandtech could get us some juicy commentary as to WHY Rocket Lake-S performed so badly, just from one of those retail samples, even without support from Intel. The opportunity was there. The NDA would not be violated. Intel wasn't lifting a finger to stop anyone from doing so. Most importantly, Anandtech needs those clicks. Intel hasn't exactly been delivering on the hype train lately. Which review articles do you think generated the most ad revenue for Anandtech (or any other ad-driven review site)? Comet Lake? Vermeer? Matisse? The 9900KS?
If or when Rocket Lake proves to be a dud of a product, how many people do you think will slavishly devote hours of their time to going through multiple Rocket Lake articles to learn everything they can about the chip? I'm thinking not very many. Between the 9900KS, Comet Lake, and now Rocket Lake, Intel is giving very little in the way of fresh meat to sites like Anandtech. Just check out the buzz around the Comet Lake launch to see what I mean. Mostly you had a bunch of people arguing over game performance and/or power draw. Now contrast that with the massive Matisse and Vermeer launch thread(s). Hype train central. People were stoked for those chips. People are STILL stoked for Vermeer, as in: when can we buy one? Please sir, I want some more. Insert Oliver Twist reference.
Intel launches aren't really generating a lot of clicks for AT (or other review sites) outside of maybe Tiger Lake, and odds are good that Dr. Cutress could see the same thing happening with Rocket Lake. Launching an early review of these fortuitously leaked retail samples would be the perfect way to scoop the scoopers and drive some badly-wanted/needed site traffic during an otherwise-lukewarm product launch. It's that or wait for Alder Lake and Warhol/Raphael/whatever AMD launches next.