There has been a lot of ink spilled, physical and digital, over the apparent connection between a Russian data center going down and and the apparent evaporation of Trump’s yahoo chorus of likes and comments. But is there any reason to assume there is anything unique about either the center or the company operating it?

By now, you have all certainly seen the memes pitying the poor MAGA chumps who suddenly lost their online girlfriends/boyfriends after Ukrainian forces took down a key data center owned by Russian search engine / AI company Yandex, as well as a slew of articles arguing that it didn’t really happen. While these articles make a reasonable argument that not all bot traffic serves nefarious purposes, they also overlook a key point – one easily overlooked by journalists who cover a wide range of ‘digital’ topics from the viewpoint of a reasonably well-informed end user.

I am not that journalist.

To the extent I have any business calling myself a ‘journalist’ at all (I prefer the term ‘essayist’), I write from a perspective that is mercifully oblivious to video games and dating apps – and one that takes a highly skeptical view of what presently passes for ‘digital culture’.

On the other hand, it is a perspective shaped by decades of experience managing digital assets and digital infrastructure – you know, all that boring nuts and bolts stuff that makes current web technology work in the first goddam place. From my perspective, there is nothing at all surprising about a correlation between taking out a data center owned by “Russia’s Google” and a drop in spurious web traffic. What does surprise me is that no one is taking this as a clear indication that a closer look needs to be taken at the other Google… as well as all the other companies that operate in this space.

Let’s start with a better understanding of how companies like Cloudflare identify ‘bot traffic’ in the first place – and why a drop in such traffic is only correlative, not definitive, evidence. The reality is that only an assumption of good faith on the part of companies like Google and Yandex enables anyone to identify such traffic in the first place. Bot traffic is identified on the basis of what is called a ‘user agent’ value encoded in all http requests. Requests lacking such a value are usually rejected out of hand by properly secured web servers, for good reason: that’s how you open yourself up to a denial of service attack. 

Since companies like Google and Yandex depend on bot-based indexing to produce search engine results, they make sure that their indexing requests include a user agent string that effectively says “hi, I’m an indexing request from some fucking search engine”. Because web sites want to be indexed (and therefore findable by end users), such requests are permitted. They are also identified as ‘bot traffic’ by network monitoring and security services (Cloudflare, among others).

But who the hell ever said that MAGA’s imaginary boyfriends and girlfriends actually identify themselves as bots?

If you are a human reading this essay, you are doing so from a web browser running on either a desktop computer or (far more likely) some handheld device. The user agent string in your web request includes that information, which is then used by code on the web server to make sure you are having the web experience you expect – and when that doesn’t happen, someone like me gets to hear about it.

The problem is that Yandex, Google, Meta, and everyone else whose business depends on operating a data center is now offering an AI enhanced browser that identifies as a human end user – even if that request is at least partially generated by an AI. It is a short jump from ‘partially’ to ‘completely’… that is, if you have a data center at your disposal that can run thousands of virtual machines running thousands of browsers pretending to be operated by something that reasonably resembles an actual human being.

This is the real problem, and it is the one that Big Tech – American, Russian, Chinese, or whatever – wants you to ignore. It is one that the European Union is trying to address, but unless a firewall gets built for the entire continent… it is not a problem they can address on their own.

There is no difference between Google acquiescing to Trump’s absurdist rebranding of Lake Ontario/Gulf of Mexico (or, god help us, something bigger) and Yandex acquiescing to Vladimir Putin’s efforts to disrupt American elections on behalf of his puppet – you know, the one Google is all too happy to appease. In both cases, the power being exercised on behalf of dictators and would-be dictators is too great to be trusted in the absence of stronger restraints than mere user agent detection and bot traffic throttling.

For now, we can trust Volodymyr Zelenskyy to ‘regulate’ Yandex. If companies like Google cannot be brought under regulatory control more stringent than having their products rebranded as ‘superintelligence’ by the same self-proclaimed stable genius who thinks he can rebrand bodies of water at whim… don’t be surprised if non-state actors decide that what’s good enough for Yandex is good enough for Google. Drones don’t care whose logo is on the side of the data center.

And neither do I.


Photo by Zulfugar Karimov on Unsplash