it’s weird (perhaps convenient) that their explanation is “bots” but they don’t mention bot metrics from StatCounter themselves: according to StatCounter numbers, bot traffic share increased 0.37% in July 2026, while Linux Desktop grew 1.92% in the same month.
Ignoring that the default browsers on many distros are seen as bot traffic, filtering the alleged bot traffic on Linux still results in around 7% of overall traffic, up from 4ish a year ago. It also doesn’t explain the giant leaps in downloads on many distros that correlate with the spike. It’s possible some organization set up a ton of computers with openclaw and a wide variety of distros on it I guess to influence opinion 🤷♀️
Also extremely biased source
I’m on pro Linux side and still think that trend was pushed by AI scrappers.
Okay, so did they account for the bots that claim to be Windows PCs? (I’ll bet there are still some running around that identify themselves as IE 6, although those may finally no longer be the majority of all bots.)
Just note how the guy starts with “only Microsoft numbers can be trusted” is biased bullshit. Pretending a company will give true numbers at all time, is a big lie. If anything they will inflate numbers and they’d try to hide the truth as much as they can because investors are watching.
While there will be truth in bots taking a part of the traffic, the whole article is Microsoft bootlicker talk.
Taking another random tool that seems to filter bot for results is just him trying to get a confirmation for his bias.
Yes a 10% stat at a given time by a given measurement tool, is not equal to a true 10% userbase. Yes there are bots. Yes there are Linux and Windows zombies alike.
Still, linux desktop is going up, with a fraction of the marketing power Windows and Apple have.
Just note how the guy starts with “only Microsoft numbers can be trusted” is biased bullshit.
I would argue misquoting someone to fit your argument is more a egregious example of “biased bullshit”.
Only Microsoft or OEM partners can tell you with the highest accuracy how many people use Windows now versus five years ago.
This is literally true. They’re not saying they WILL tell you the accurate numbers, they’re saying only they have those numbers to tell.
To be fair…
This is ‘windowslatest.com’.
Totally agree with you, but… the bias … was foreshadowed, lol.
Still, linux desktop is going up, with a fraction of the marketing power Windows and Apple have.
This isn’t entirely true. Linux is getting a lot of help from Microsoft. <wink wink>
FWIW this website is always like this, every article I’ve read from latest feels like it came from the MS PR desk.
I wonder why that could be, rofl!
Surely MSFT would always be clear, transparent, and forthright with the public, ahahahaha!
Trust that Microsoft will give true numbers, wink wink
No offence to the artist but the WSL Tux is ugly af
they deserve the hate
If it were me I’d have gone with
but with penguin instead of cat(Windows get it?)
What a sicko.
Copium
To address some of the author’s points.
- MacOS vs OSX: to my understanding, safari reports the OS as OSX still, likely for compatibility reasons. 3rd party browsers such as chrome changed their agent to macOS, which explains the split. macOS users are more likely to use Safari in general.
- Bots: Does the author really believe these sites aren’t capable of detecting bots? I’m sure some slip through, but if not traffic were included in general, then Linux would be at a 90% share or higher. They just consume way more page traffic than a human ever could.
- Data corrections: Sure, the data may get corrected a bit to bring Linux numbers back down, but there is a clear trend that Linux numbers are rising. It also happens that Windows is where people are migrating from the most.
- Official numbers from Microsoft: lol
Bots: Does the author really believe these sites aren’t capable of detecting bots? I’m sure some slip through, but if not traffic were included in general, then Linux would be at a 90% share or higher. They just consume way more page traffic than a human ever could.
Only other possibility I can think of is llms made botting for small operators more accessible. Still leads to a lot more people running linux
Yes there are zero bots running windows and nothing would ever inflate those numbers pleased clap.
There are thousands of organizations that rely on Microsoft for their office suite, Outlook, Teams and these companies have thousands of PCs. They can mandate what update will get pushed, restrict admin access, running monitoring tools etc… These are hard to replace.
Overall probably closer to the truth.
I wish people didn’t obsess over these large numbers as much. Like when Linux suddenly hit 5% on Steam when that was clearly just an outlier and was destined to fall right back down.
The only interesting thing about the numbers is that eventually, people outside the tech bubble will pick up on it and think about seeing what the fuss is about and if it’s a solution for them and then maybe also switching.
In that sense: I “heard” that the number is 20% actually. Windows is so over. It’s true. Pass it on.
How would that clearly be an outlier? Hindsight is 20/20, right?
Because these numbers experience wild fluctuations on a regular basis. Does it make sense to you that people are just joining and quitting Windows by the masses simultaneously?
Something strange with the March 2026 results was that it had two new distro entries, “0 64 bit” at 17.60% and “64 bit” at 8%. In April, one of those entries disappeared and the other’s share dropped significantly. In May, both were gone entirely and Linux was back to a more reasonable marketshare of 4%.
I’ll admit that weird and unexplained data is a hint to something going wrong. Shame on steam.
There was nothing that would have explained a more than doubling of Linux desktop market share.
It was not that difficult to come up with a more likely explanations and some people where quick to mention AI crawlers/agents, which did actually explode in numbers.
But why would those crawlers participate in the steam survey?
The latest results I saw from steam actually showed an increase in Windows 11 market share.
The 10% was based on browser user agents.
guys don’t tell them that this also means that windows is poop, since it’s either too bad at scaling or at tech infrastructure to even attract ai developers to the platform 😂
I’m betting that the bots are running in something like ubuntu-node docker images which could’ve been built from any OS.
I’m also willing to bet that the survey was only ever looking at the OS reported in the User-Agent which those scrapers will change frequently.
Who cares, just fake it till you make it.
Humans are why we cannot have nice things. The internet was always going to get ridiculous, just like the usa was always going to prove back to the future right in the stupidest possible way.
pcmasterchef
