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.
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).
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.
API protocol says “return value is either “OK” or “ERR””, attacker sends “Floccinaucinihilipilification”, overflowing the 3-byte response buffer allocated on the stack.
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.
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.
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.
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.
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.
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.
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.
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.
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? -_-
>It's a built-in tool of the OS, not some random spyware
Oh no dont get me wrong, it's definitely not some random spyware, it's deliberately made by Microsoft targeted at Windows Users who cant buy a Enterprise License.
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?
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.
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.
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."
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.
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.
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].
You can't have finished software any more. It's now illegal or something. Must keep changing things forever. And must add some AI features of course, to make the VCs happy.
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.
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"
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.
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.
- 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.
Presumably as part of the generative AI stuff they crammed into it? They've been hilariously cagey about the details here which makes me think it has to be AI-feature related
Not sure what you think is "hilariously cagey", but why would they only be "cagey" if the bug was from ai-features?
Also it says paint versions are "affected from 11.0.0.0 before 11.2605.91.0" which I take to mean the bug has been there since Windows 11 was released in 2021
> 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.
At the moment I just use Zed, or I imagine I'd be fine with Sublime. When I used Jetbrains/VSCode, I would grab Notepad/Notepad++ when I didn't want to wait for them to open, set up a project, etc. But if something 'heavyweight' is quicker than notepad, and the features don't get in my way, I don't need anything else. In fact I still appreciate the extra functionality with those one-off text/data/scripts - format, regex replace, diff, occasionally even LSP like shellcheck
mspaint.exe is quite likely to have at least one code execution vulnerability if one opens a malicious file, but I bet it's very safe against remote attack :)
> 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.
Firearms are now exposing the inherent vulnerability of human flesh against small projectiles flying faster than the speed of sound. Before that we were just fist fighting eachother and we were trained for that. But no mo! Get your dragon's breath today at $199.99!
I was saying this is just one case where we can say the reason about copilot, there are dozens of reported vulnerabilities, bugs in kernel, rust. and many more elite products which were authored by so called elite developers.
You willingly let "perfect" be the enemy of "good" in this case. It's absolutely possible to build software runtimes that obviate this type of nonsense. It's not grandstanding by elite developers, most of these choices are common-sense mitigations.
--Another vulnerability in ssh accounts that should never take a password at all (root, your key-only deploy user, ldap users, the default user on alpine's aws and azure images) take any password longer than 256 characters.
I guess these also belongs to your common sense mitigation category but this has been there since 2016.
-- struct's lifetime requirements can be bypassed via covariant field assignment,
11 year old rust compiler bug. how about this?
Why on Earth do you have to give Paint, internet? How is that a technical debt? Making all nice little fully offline apps Internet Nightmare Slop is wag beyond debt.
It might not be paint with internet, it might be something like a packet over your network asking a computer to open an input/file, and the computer responds by opening that malicious input with paint
Presumably this is like many other graphics tool vulnerabilities this works by the attacker persuading a user to open a file that exercises the bug. Not terribly unusual.
I'm no expert but so much of Windows is implemented as services. svchost.exe is probably involved somehow. Like it wouldn't surprise me if Copilot is implemented as a service and it talks to Paint via IPC messages or something
It would make sense that it could listen for messages like "your generated image is ready!" and pass bitmap data to Paint. And media like images are a common source of vulnerabilities since you're basically passing binary data into the software.
This site absolutely does jokes, it just doesn't beat horses into bloody pulps with the jokes the way reddit does. We have a higher standard. A joke has to actually be funny, and not some tired reference.
Not to imply any particular comparison to less serious websites where the karma system over time may have resulted in a disproportionate surplus of one liner jokes and other spammy low quality posts.
Tangential, but: "We're sorry but the CVE Website doesn't work properly without JavaScript enabled."
Usually I just enable JS. But is this a security site, asking for remote code execution on my device? Before they can show me some text? They just disqualified themselves.
I'm just in full denial that paint runs a server. Happy to be corrected if someone has a windows machine handy to verify.
But From [0]:
> A total of 59 of the October CVEs fit my definition of wormable. Specifically, 57 contain both “An unauthenticated attacker could exploit this vulnerability by sending a specially crafted packet to an affected service over the network” and “No authentication or user interaction is required.”
Makes you wonder if MS is just copy-pasting boilerplate on CVEs and disclosed incorrectly the scope of the one in Paint because they had 56 other bugs to disclose.
So I'm not discounting reporting human error here, yet.
"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.."
It mentions that it may rely on other services or security measures running- I wonder if there’s something like this hitting a firewall or av and then that calling out to paint to check file type or size out something like that?
But why MS Paint? What functionality does it offer to another program to get called on specifically? And if this other service isn't calling on it specifically, what else can it be tricked into calling? And how many of those are also vulnerable to being stabbed in the back? The smaller interior holes are a problem, sure, but it seems like the CVE should be for the hull hole in that case.
Sounds like they just mean opening a file that you got from somewhere else.
>The CVSS score of 8.1 marks this as a high‑severity RCE. The EPSS score is not available, and the vulnerability is not listed in CISA’s KEV catalog. Based on the description, the likely attack vector is remote: an attacker can deliver the malicious file over a network connection, for example via Remote Desktop, file shares, or other network‑exposed services that allow users to open images. The lack of EPSS data does not diminish the threat posed by a known RCE, but it does make precise likelihood estimation difficult.
> After applying the update, verify that the target Windows systems are running a version that includes the CVE fix before disabling legacy file‑open processing in Paint.
Which is complete nonsense. "Legacy file-open processing" isn't a real configurable option in Paint and this seems to suggest that enabling it is some sort of mitigation. The resolution is simply to update Paint.
The direct implication is that current generations of ms paint are a “service over the network.” Probably a theoretically possible RPC end point of svchost.exe
I'm thankfully out of the loop on Windows these days, but as I remember it Paint did not have any network features? Is that still true? If if is, maybe it is a misfiling of the CVE and it isn't remote?
Or are they chaining a browser auto opening a download in paint, (but browsers shouldn't do that)...
Why for the love of everything that is good does MS Paint have network trash in it. It's a super simple bitmap editor. Like Notepad, it's simple and just works.
Make a new product and bundle that as an option if you want to drastically change the functionality.
> Notepad in Windows includes AI features powered by Copilot that help refine and shorten text with the assistance of GPT. By default, these AI features are enabled on managed devices
I wonder if it could be through a file:// or some deeplink custom url handler that opens in Paint and clicking on it with some carefully crafted file path could make it read out of bounds (since that's the associated weakness reported) and execute something.
Whatever it is, it doesn't require any clicking or user interaction. Maybe my imagination is just limited but I can't imagine why Paint is making itself available for incoming traffic on the network.
Mine doesn't have a mouse or keyboard at the moment, so I'm only checking over an SSH->WSL connection. But it sure doesn't look like it's opening any ports from here.
So I guess either some random OS component is listening to the network and just-as-inexplicably asking Paint to do something for it. Or one of Paint's outbound connections is vulnerable if hijacked by a mitm. Network interception's one of the examples used for the "High" Attack Complexity rating that MS gave this CVE. Maybe powered by eMudhra for any systems that haven't revoked those certs? That'd be topical.
But still wouldn't really make sense. A patch for machines that are actively updating Paint but dodging cert revocations? Nothing really makes sense about this silly Paint CVE.
Or maybe after the .bmp file is executed, it can install something that opens an outbound TCP connection to its C2, and thus can receive commands. Would not need admin privileges because it doesn't open a port, just 'outbound' connections, like a long-poll.
you are right, it seems like the latter outbouns tcp connection is already being opened by paint, and that is what is exploitable.
since it exposes no port, I’d assume it isn’t trivial to exploit, paint will chose its target (a microsoft domain or IP), the exploit must involve some error in destination selection, or simply allowing a mitm to impersonate microsoft.
this is inline with Attack Complexity High, attacker needs to be either in the network path (mitm) or control the erroneous destination of the connection (say mircosoft.com), possibly blocking (dosing) the higher priority genuine destination
Seems odd even for a vibecoded copilot integration to make repeated, unprompted network requests. And if they're not repeated and unprompted, then we don't meet the "no user input" requirement.
Because the definition of "remote" has been expanded in the name of resume-driven development to include sending a malicious file that a victim has to open.
> The vulnerable component is bound to the network stack and the set of possible attackers extends beyond the other options listed, up to and including the entire Internet.
The way I read it is MSPaint is now running as a service and listens to a port (while running? all the time?) and you can trigger remote execute by sending data. Just.. Why?..
At this point I want to declare that air-gapped network with the transfer only over USB drives is a viable and valuable method of handling IT security. Outside transfer only happens by people manually copying text onto a floppy and reading it from a floppy.
I love Microsoft! Who else might look at a fully offline drawing app and think: "Man this needs the most severe level of networking access capable of remote code execution", and that surely is to put their useless Copilot to absolutely nowhere. Next headline will be "All user credentials stolen from Notepad"
We seriously need more information.
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).
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.
API protocol says “return value is either “OK” or “ERR””, attacker sends “Floccinaucinihilipilification”, overflowing the 3-byte response buffer allocated on the stack.
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.
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.
It's quite common with vulnerabilities found by IA, there might be less hallucination, but the impact/score is often completely wrong.
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.
Then they added automatic HTML parsing of incoming emails, and the attackers cheered.
You can go back to mutt's parsing, and CVE-1999-0940 for an old example. But I'm sure that's only one of many.
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.
And at the end of the day $2 and $2000 screwdrivers are both still screwdrivers.
once Microsoft has reinvented themselves as…security-first?
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.
https://interestingengineering.com/culture/rip-microsoft-pai...
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...
If I still mamaged any windows host, i would be deploying a remove-package script asap to the entire fleet.
When did y'all stop using some basic sense before writing? -_-
Yeah samesame:
-Compares a kernel with paint
-Uses a program with telemetry/spyware instead of installing on of the 1000s alternatives
Oh no dont get me wrong, it's definitely not some random spyware, it's deliberately made by Microsoft targeted at Windows Users who cant buy a Enterprise License.
Now that I'm on Linux, I use jspaint.app.
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.
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.
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.
[^1]: https://lawsofsoftwareengineering.com/laws/zawinskis-law/
[^2]: https://www.jwz.org/hacks/
https://hn.algolia.com/?q=mspaint+ide
>june 2026
Pretty sure it would be May?
But in either case they had 91 releases that month?
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.
Are you using Windows? If so, then yes.
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"
This is some Eric-Cantona-level gnomic mysticism. I approve.
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
Could also be some kind of recall feature.
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.
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.
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... . . .
> 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...
Good for you.
Now only <all Windows users>-1 are affected.
[0]: https://liquidninja.com/metapad/download.html
"No User Interaction"
Does that mean that paint doesnt have to be running?
Application being installed on a PC allows a specially crafted packet to arbitrarily execute code?
0. My gut says this is somehow related to indexing/thumbnail rendering. But maybe I'm off base.
1. The 'no user interaction' could be related to (ironically) some background pre-indexing process that windows is doing on files inserted to folders.
1.1. LOL maybe their bullshit 'Windows Copilot' which is why the description is so vague, don't want to torpedo things worse than they already are.
2. It could also be, still in the category of 'well, ok they click on a share folder and the preview runs' or a weird edge thing.
2.1. But they aren't being that clear yet, probably because it's bad.
UI:N it is with no interaction ?
is it a cve mistake ?
> 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.
Until somebody connects the dots on how the heck MS Paint is accessible over the network, I'm gonna maintain some pretty intense skepticism here.
> https://share.google/aDP2ac9j5AmB13QRo
>Unproven
>Remediation Level
>Official Fix
>Report Confidence
>Confirmed
I really like these secondary scoring factors, not just scoring the vuln itself, but how it's being attacked.
My take is that the addition of CoPilot to MS Paint is probably how this CVE got introduced.
> LLMs are now unfolding technical debt, poorly written code, and security vulnerabilities accumulated in software that they have created themselves.
https://msrc.microsoft.com/update-guide/vulnerability/CVE-20...
> User Interaction: None
> Out-of-bounds read in Windows Paint allows an unauthorized attacker to execute code over a network.
What server is MS Paint running or what services does it provide?
Edit: Apparently you can use it to print an image from the command-line. Is that how this is being exploited?
Except for the infamous Dropbox comment.
Usually I just enable JS. But is this a security site, asking for remote code execution on my device? Before they can show me some text? They just disqualified themselves.
But From [0]: > A total of 59 of the October CVEs fit my definition of wormable. Specifically, 57 contain both “An unauthenticated attacker could exploit this vulnerability by sending a specially crafted packet to an affected service over the network” and “No authentication or user interaction is required.”
Makes you wonder if MS is just copy-pasting boilerplate on CVEs and disclosed incorrectly the scope of the one in Paint because they had 56 other bugs to disclose.
So I'm not discounting reporting human error here, yet.
[0] https://www.zerodayinitiative.com/blog/2026/10/13/the-octobe...
But remote? How
Still, the question remains... How?
>The CVSS score of 8.1 marks this as a high‑severity RCE. The EPSS score is not available, and the vulnerability is not listed in CISA’s KEV catalog. Based on the description, the likely attack vector is remote: an attacker can deliver the malicious file over a network connection, for example via Remote Desktop, file shares, or other network‑exposed services that allow users to open images. The lack of EPSS data does not diminish the threat posed by a known RCE, but it does make precise likelihood estimation difficult.
https://app.opencve.io/cve/CVE-2026-83844
> After applying the update, verify that the target Windows systems are running a version that includes the CVE fix before disabling legacy file‑open processing in Paint.
Which is complete nonsense. "Legacy file-open processing" isn't a real configurable option in Paint and this seems to suggest that enabling it is some sort of mitigation. The resolution is simply to update Paint.
Or are they chaining a browser auto opening a download in paint, (but browsers shouldn't do that)...
Make a new product and bundle that as an option if you want to drastically change the functionality.
> Notepad in Windows includes AI features powered by Copilot that help refine and shorten text with the assistance of GPT. By default, these AI features are enabled on managed devices
https://learn.microsoft.com/en-us/windows/client-management/...
I mean… what else were the people working on Notepad and Paint to do!?
The order came down to integrate AI, and they wanted to be promoted.
Every single day I’m happier and happier than I use Linux now.
Should be easy to verify for someone with a windows machine if it does.
So I guess either some random OS component is listening to the network and just-as-inexplicably asking Paint to do something for it. Or one of Paint's outbound connections is vulnerable if hijacked by a mitm. Network interception's one of the examples used for the "High" Attack Complexity rating that MS gave this CVE. Maybe powered by eMudhra for any systems that haven't revoked those certs? That'd be topical.
But still wouldn't really make sense. A patch for machines that are actively updating Paint but dodging cert revocations? Nothing really makes sense about this silly Paint CVE.
Or maybe after the .bmp file is executed, it can install something that opens an outbound TCP connection to its C2, and thus can receive commands. Would not need admin privileges because it doesn't open a port, just 'outbound' connections, like a long-poll.
> The vulnerable system can be exploited without any interaction from any user.
since it exposes no port, I’d assume it isn’t trivial to exploit, paint will chose its target (a microsoft domain or IP), the exploit must involve some error in destination selection, or simply allowing a mitm to impersonate microsoft.
this is inline with Attack Complexity High, attacker needs to be either in the network path (mitm) or control the erroneous destination of the connection (say mircosoft.com), possibly blocking (dosing) the higher priority genuine destination
States:
> Attack Vector: Network
> The vulnerable component is bound to the network stack and the set of possible attackers extends beyond the other options listed, up to and including the entire Internet.
The way I read it is MSPaint is now running as a service and listens to a port (while running? all the time?) and you can trigger remote execute by sending data. Just.. Why?..
I just... It's just ridiculous at this point.
[0] https://x.com/orange_8361/status/2115654643141489057?a=qallN...
Each time I learn something new about Microslop, I'm just glad I'm not a tenant.
Cross hatch, Scottish tartan?
This is absurd.