MS Paint Remote Code Execution Vulnerability
cve.org709 points by kuberwastaken 2 days ago
709 points by kuberwastaken 2 days ago
Paint can exectue arbitrary code from the REMOTE, without USER INTERACTION?????
We seriously need more information.
Paint’s image decoding engine is a shared system component and during thumbnail and visual preview generation it gets called automatically without opening the GUI. Ergo it’s enough for a malformed image file to exist on your device to get pwnd.
Do you have details on this? Given that this is classified as a network vulnerability of the Paint application itself rather than any shared system components, I assumed it had to do with the recently-introduced Copilot features which AFAIK are the only part of the application that directly interact with the network.
Every program which has telemetry can. And web browsers.
While I understand your phrasing getting downvoted, I agree with what I assume you mean:
1. Telemetry as implemented is usually sloppy. Even though intuitively one could think that telemetry means “data gets collected locally and fired off to a remote server”, we have seen time and time again that they also build a return path to make the telemetry gathering more powerful. That return path can be a vector.
2. Web browsers have a _huge_ surface and they connect to remote servers constantly. The industry does very well at being proactive and at addressing CVEs but the whole premise makes that very difficult and so we cannot absolutely claim that browsers don’t have RCE(s).
> That return path can be a vector
What return path could possibly lead to a RCE? That's crazy, no telemetry requires "backend could return any command that gets run on the client", you're doing telemetry in a very non-standard way if this is even possible.
At worst, the telemetry backend might first send the client a config, but none of that would contain executable commands because again, that'd be crazy.
> we cannot absolutely claim that browsers don’t have RCE(s).
Of course, you can never prove the absence of bugs (and/or security issues), but given that browsers aren't wildly exploited at this very moment, nor have been for quite some time, we can pretty confidently claim there are no browser RCEs in wide use today.
It's true that browsers have pretty big attack surface, but it's also true that it's the class of software that gets the more pentesting, since the payoff would be so great when you manage to find holes.
> What return path could possibly lead to a RCE?
API protocol says “return value is either “OK” or “ERR””, attacker sends “Floccinaucinihilipilification”, overflowing the 3-byte response buffer allocated on the stack.
Right, so you have the first step of something, but then depending on everything else, this is either a DOS or a RCE.
I'm claiming if "everything else" does indeed lead to a RCE, you're again doing something very wrong here. It's not rocket science exactly.
It’s harder nowadays, with address space layout randomization and no-execute stacks, but the basics are
- https://en.wikipedia.org/wiki/Stack_buffer_overflow#stack_sm...
- https://en.wikipedia.org/wiki/Return-oriented_programming
I think that’s how lots of RCE bugs start, so it may not be rocket science, but lots of software falls for it.
I'm sorry, but I still find it hard to believe you work with other professionals that when they implement telemetry in client applications, that usually ends up exposing a RCE on the client because of the telemetry implementation.
> Every program which has telemetry can
I'd go out on a limb here perhaps, but I'd claim that if your telemetry means that the one SENDING the data gets a RCE by default by it, you're doing telemetry very wrong. Maybe look into different libraries/services than the ones you use today, please.
If I had to guess, without diffing old/new, it's probably related to the generative erase/ai shit.
To stack assumptions, I'd reach for something like using the recent certificate circus to present as there being an updated version of whatever model it uses on the PCs that can run it locally to give it a fucked up model that does the out of bounds shit as part of whatever verification/checks it does after its downloaded/installed.
> If I had to guess, ... > To stack assumptions, ...
So, you're saying nothing? It would have been more useful to never have posted.
Either the description or the impact is wrong. You can't trigger MS Paint in a '0-click' attack, and OOB read are not sufficient to directly obtain code execution. So I'll let you guess which one is wrong.
It's quite common with vulnerabilities found by IA, there might be less hallucination, but the impact/score is often completely wrong.
I can't wait for the Raymond Chen write up one day.
Just after I deployed a "currently patched" system yesterday and failed to uninstall Paint when I had the chance.
Normally have been 86-ing a number of annoying things such as OneDrive, but leaving Paint in place figuring it's got to still be harmless.
Creeping increases in attack vector when just the opposite is needed more than ever.
Some years ago, I ended up on a call with a support team guy, helping him under some parts of a system I worked on. He shared his screen, started drawing diagrams. It took me a whole 10 minutes to realise it was in paint, he was just that good at using it.
Turns out paint was the thing almost every customer had installed/had access to, so he worked to the lowest common denominator and got really good at using it. I guess now he'll need to find a new tool to astonish devs with.
If you're in support across multiple customers, especially in regulated industries, you learn to rely on the minimum common denominator, which usually means OS utils.
And at the end of the day $2 and $2000 screwdrivers are both still screwdrivers.
This is also why knowing the basics of "vi" is so useful on Linux/unix systems
...especially knowing how to save/exit without saving/quit, when it automatically opens up for some reason.
Before MS, there was a time where it wasn’t possible to get a virus by receiving an email. Users had to click on the attachment. Email itself is a text protocol after all.
Then they added automatic HTML parsing of incoming emails, and the attackers cheered.
86-ing?
It comes from working in kitchens. It means to remove something from the menu, or otherwise make it unavailable.
Traditionally tossing excess baggage overboard when the situation gets grim enough.
Like when a previously sturdy old sailing ship starts taking on water, for instance in the 18th century and beyond.
Numerical equivalent to "deep-six."
Perhaps someone else,
once Microsoft has reinvented themselves as…security-first?
They supposedly did, that is why most stuff on Azure is now a mix of Java, C#, Go and Rust with very little C and C++.
And there is the SFI program.
However this is a Windows application, and in spite some Rust advances in WinDev, it is pretty much C and C++ turf, with Rust getting the same warm reception as .NET used to enjoy.
Guess which language is Paint written on?
Nonetheless, they are doubling down on sandboxing Windows, macOS, iOS and Android style, for quite some time, exactly to become security first as you are asking for.
I didn't know "just don't bloody add an internet access to everything" required a different language
Not asking - just saying aloud that they seem to feel the need to get ahead of sentiment (like being dev-unfriendly), and surely this one comes next.
Had to go to Microsoft Store to get the fixed version, Windows Update didn't show anything.
Windows Update and Store are slowly becoming the same, and Paint gets delivered via the store for quite some time, there was even some drama related to it.
https://interestingengineering.com/culture/rip-microsoft-pai...
Ugh so I have to unblock the Store in my firewall just to get patched. What a PITA.
Did the same and it quickly updated itself when I went into the "Paint" page (but I had to go there first). No indication of what just happened a few seconds later.
I did the same, then ran:
Get-AppxPackage -Name Microsoft.Paint | Select-Object Name, Version
But I can only get Microsoft.Paint 11.2609.401.0 and Microsoft has said different things in different places, so it's not even clear if this has the fix or not.
This article has more info on the confusion they're causing: https://windowsforum.com/news/cve-2026-83844-windows-11-pain...
why get package after something that is obviously a malicious backdoor?
If I still mamaged any windows host, i would be deploying a remove-package script asap to the entire fleet.
Get-AppxPackage gives you the currently-installed packages for a Windows profile in PowerShell. It does not, in fact, download or install anything.
The whole command I gave gives you the currently installed version of MS Paint so you can see if you're already vulnerable or not.
What kind of nonsense are you writing here right now? When the next Linux kernel CVE happens, you'll also delete all the kernels from your fleete because they're a malicious backdoor?
When did y'all stop using some basic sense before writing? -_-
Maybe we can blame PowerShell a bit for confusing names on this one. Get-AppxPackage does not download or install anything, contrary to what one might wrongly assume.
If MS knew it would drive people to actually use their app store, they would have created the CVE themselves.
For a split second, I visualized you walking into a brick and mortar Microsoft Store to get your security update done on your laptop, and somewhere in the middle of that split second I caught myself thinking „typical Microsoft“, and then I came to my senses.
This seems to affect only paint 11.0.0.0 up to, but not including, 11.2605.91.0. For folks like me, running a much older version, its probably safe, although its Windows so, I'm not too sure.
I'm amazed MS Paint is somehow up to version 11, let alone 11.2605 (!). I literally can't remember the last time I used it, but what are they even changing?
They've added a lot of features and it's almost like Photoshop Elements now. It has layers, general photo editing tools, an AI generator, etc.
I remember that when Paint started being funky I found a Windows XP exe and used that. I want Paint to be the graphical equivalent of hexedit.
Now that I'm on Linux, I use jspaint.app.
I still use the Paint from Windows XP, running in Wine.
I would use kolourpaint but for whatever reason overlaid GUI elements like the selection box and brush cursors leave trails on the canvas.
Fun anecdote, paint.exe crashed the other day just as I was finishing up my drawing of the tower of Babel.
Honestly I like layers in Paint. Even pixel art programs have layers.
I remember when I was doing pixel art, there were some purists that insisted that using layers and transparency/opacity to create windows for your buildings is more or less cheating. According to them, one should manually figure out what the color palette should be for things seen through a window.
Anyway, I continued using layers and transparency/opacity. Pixel art was hard enough already for me, so I didn’t feel that it made sense to artificially make it even more difficult for myself at the time. I had enough other aspects of pixel art to work out already in order to try to get better at it.
XP era mspaint with layers is really all I want. Rudimentary image editing and drawing but with layers.
They should have just hired the paint.net guy and rolled it in.
The point of paint is a tiny pixel editor. If you want layers there are better programs.
But is Windows 11 paint capable of pixel art? I'm reading that it antialiases everything.
You can draw single pixels. I'm not sure where whoever told you that got it from.
And create single-pixel images. My most used application of Paint by far is creating single-pixel PNG files to work around the fact that "solid color" is not an option for lock/login screen backgrounds.
But can you resize without antialiasing, and does pasting trigger antialiasing?
There's this https://www.windowsphoneinfo.com/threads/why-does-resizing-m... (made confusing by the next person talking about Paint 3D, a different application) and this https://learn.microsoft.com/en-us/answers/questions/3991946/... (made confusing by the person being an artist and talking in peanut butter metaphors).
There's some suggestion there that shapes (including lines) are separate entities from pixels. Wikipedia references a now-deleted answers.microsoft.com version of the first question https://web.archive.org/web/20250203164656/https://answers.m... with a couple of people saying that resize is useless. Wiki also says "re-copied" images blur, which is why I wondered if that meant pasting. Also this: https://learn.microsoft.com/en-us/answers/questions/5673783/... a complaint that the tool for drawing pixels doesn't indicate exactly which pixel it will draw to.
I can't reproduce any of those issues. When drawing the indicator snaps to a pixel, so it's perfectly clear where it's going. When I draw a line, it's pixel-perfect when moving the control points, likewise after it finalises. And when I copy a section of the image and paste it back in, it's pixel-perfect as I move and when I finalise. There doesn't seem to be a true resize; all I seem to be able to do is change the size of the canvas, which doesn't add any kind of aliasing.
But I've only used the program for less than 5 minutes, so maybe it's configured differently?
Yes, resizing within Paint preserves pixel-level color. I just resized a single pixel dot to about 14x25 without aliasing.
I don't know why you are so concerned, but it's incredibly easy to test these questions if you really want to know.
Just like any text editor is doomed to become either Emacs or Word, so any image editor is doomed to become Photoshop.
Reminds me of Zawinski's Law of Software Envelopment[^1,2]: "Every program attempts to expand until it can read mail. Those programs which cannot so expand are replaced by ones which can."
[^1]: https://lawsofsoftwareengineering.com/laws/zawinskis-law/
Too true. I gave up trying to fight it years ago and now write all my documents in Emacs and edit my code with Microsoft Word.
I hope MS is not getting exclusive right for manufacturing spoons because I want mine to scoop up stuff like soup or ice-cream without smart features activated by elegant touch UI on the handle like cutting down trees being in the way to the grocery store to 'elevate user experience', or others. This stuffing more and more into everything and anything for the sake of shock and awe is a stupid (and dangerous) thing.
It's the last two digits of the year plus the month number. That was June 2026. It was 2504 in May 2025, and 2402 in March 2024.
>2605
>june 2026
Pretty sure it would be May?
But in either case they had 91 releases that month?
You haven't spotted that it's regularly being released the month following the revision number. (-: Because Microsoft will of course have a release pipeline that things have to go through.
They're not a positional notation. It's version 11.2605.91 recently. It was 11.2605.81 earlier this year. There was an 11.2605.71 and an 11.2605.61 before that. Last year, there was a 11.2504.451 . The year before there was a 11.2402.32 .
Microsoft's versioning systems are not 4 columns where a tick around of one increments the one before it. The easiest way to think of them is madness made for marketing purposes. (Windows NT starting at 3.1, for example.) There is some logic to it, but nowadays it's often four entirely different things jammed together, ticking around independently. Like the version of Windows, the year and month of the release, and perhaps a daily build number or some such.
Windows NT itself has a monotonically increasing build number as part of its versioning that started some time in 1992 and is currently just over 26 thousand.
The Windows 11 version is a complete rewrite.
So was the Windows 7 version
Windows 7 added ribbons to the UI. It's unlikely it was a rewrite. Just a retooling of what Vista had, which in turn what Windows XP had, which... [continue down the line until Windows 1.0].
This was my observation too; I assume 11.0.0.0 is windows 11? because my windows 10 mspaint doesn't even match that numbering format.
yes that is my exact situation. I think (hope?) Microsoft would've disclosed it if prior versions were affected too, although I never heard the "old" paint have anything to do with network, so there's that.
My version shows "22H2" I don't know if I'm vulnerable or not
That is Windows version. But you probably use the regular Win32 Paint, so should be okay.
> I don't know if I'm vulnerable or not
Are you using Windows? If so, then yes.
Or if you're using Linux. Or OSX. seL4 users are safe though, mostly because their computers don't do anything useful.
I hope I can use this to embed a functioning Piet interpreter into Paint so it can execute arbitrary programs that I draw.
WHY ON HELL paint has a remote code execution in the name of God? WHY? is this because someone who needs to build a CV decided to network something on ..... I can't come with a reason why "An unauthenticated attacker could exploit this vulnerability by sending a specially crafted packet to an affected service over the network." will cause paint to start being nice to an attacker and leaking crypto secrets etc.. the thing says sending a specially crafted packet.. this means windows is part of the attack vector, but that just makes the WHY bigger, really, why?
>I can't come with a reason why "An unauthenticated attacker could exploit this vulnerability by sending a specially crafted packet to an affected service over the network."
The Microsoft link doesn't mention this at all, so my guess it's "remote" in the sense that a zip file exploit is "remote"
It absolutely does, in the "FAQ" section:
https://msrc.microsoft.com/update-guide/vulnerability/CVE-20...
> How could an attacker exploit this vulnerability?
> An unauthenticated attacker could exploit this vulnerability by sending a specially crafted packet to an affected service over the network. Successful exploitation could allow the attacker to execute code on the target system. No authentication or user interaction is required..
Paint now has Copilot features that require sign-in, so presumably involve network access.
Put it on a file share and now any file is "remote".
Remote, and inapplicable. One of the very few things the Microsoft link does mention that no user interaction is required. A banana left out in the walkway still requires the user to go for a walk.
> A banana left out in the walkway still requires the user to go for a walk.
This is some Eric-Cantona-level gnomic mysticism. I approve.
Any extraction of behavioural surplus is a win, Nadella went hard on this in Windows 11. If you are drawing something for your child in paint, you might want to buy pampers or something.
my first reaction was "oh_ why??!?, not again_"
not sure if the last time was also a RCE specifically, but this by far isn't the first "this shouldn't even be remotely possible" CVE in MS Paint
NSA backdoor?
unlikely
IFF there is a NSA backdoor it likely would be:
- not in paint, but something you can't disable/remove in hardened special purpose windows setups
- not be easily usable by 3rd parties, like some auth accidentally accepting some quite specific other password (e.g. due to some bug reading the expected hash from the wrong place) or similar
lets not forget as much as the NSA might want to spy on you they also want to keep China/Russia out of spying on US companies. A traditional RCE in paint is very counter productive for that.
unless they actively work with MS to keep their backdoor fresh, one can reasonably assume a potential backdoor _was_ well-hidden and advanced for the time period when it was introduced. What about after 20 years and LLMs decompiling everything?
They (and other governments) can do multiple things at the same time.
One one hand they can actively try to work with companies to get secret back doors put in, sure. But I imagine that's not usually very successful. Lavabit chose to go out of business instead of doing so... I wonder who else is out there that we don't know about.
But we also have evidence that they actively look for, buy, and hoard, unknown (to the vendor) exploits for their own possible later use in case they need it for something.
I'm sure you can imagine that it's much more advantageous to keep collecting all kinds of different ways to get into a system in case one method is not applicable for any number of reasons.
I believe they're already quite happy with the Windows OS itself for this purpose, and probably some other firmware-level rootkits across many products like Intel and Qualcomm chipsets. Their contracts with Meta might be helpful too.
llm they use to vibe-code windows was supposed to make no mistakes and make everything secure...
unfortunately it made some mistakes and didn't make everything secure.
im sure windows security will get better as models get better...
maybe in six months when we have AGI... . . .
Paint opens a port?
Yes, when it launches, it opens a telnet server so that bill gates can log in and type all the colors for you to use, like #000000, #000001, etc
What Network?
It's probably a way to get phished, if you click on a malicious bmp file.
MSRC[1] says otherwise:
> An unauthenticated attacker could exploit this vulnerability by sending a specially crafted packet to an affected service over the network. Successful exploitation could allow the attacker to execute code on the target system. No authentication or user interaction is required.
[1] https://msrc.microsoft.com/update-guide/vulnerability/CVE-20...
I don't know anyone running Paint servers and I'm pretty sure phishing could be possible.
It's bizarre that you're doubling down in the face of statements from Microsoft's security experts who have the full context of the investigation and the source code. It's worth some self-reflection on what drove you to make this post.
The boilerplate FAQ text is frequently wrong or incomplete. You can go through the catalog and look through some of them and see how many have multiple revisions. I think, if you want to know definitively, you have to patch diff it
Where does it say that you can't be phished with a bmp file?
Maybe you should self-reflect on why you are offended by my comment. You might learn something about yourself.
It says that "no user interaction is required", so phishing is explicitly not what this vulnerability is about.
You haven't declared any privileged knowledge of this attack or MS Paint and have, on vibes, doubled down on what you assume must be going on despite other evidence.
It's a bad look.
Look away then.
You don't have to comment on other people's comments you know.
Phishing with .bmp files existed for a very long time now.
> Maybe you should self-reflect on why you are offended by my comment
A significant chunk of the world's problems today are caused by or exacerbated by people who confidently assert things they believe regardless of what evidence or experts say, and who dismiss expertise. Those people are emboldened by others allowing them to spew nonsense unchecked, and I have no interest in seeing HN degraded similarly.
You're moving goalposts. A commenter asked what Networking MS Paint had in the context of this vulnerability. You made an assertion not supported by the documentation we had, and now you're trying to claim that there could be other vulnerabilities rather than admitting that you made a mistake without doing any research.