This feels like a case of "It rather involved being on the other side of this airtight hatchway"[1]. If you can read arbitrary process memory, you're probably also in a position to just dump out the passwords by pretending to be the user in question.
> If an attacker gains administrative access on a terminal server, they can access the memory of all logged‑on user processes.
If an attacker has administrative access, they can also attach a debugger to every chrome process and force it to decrypt all the passwords. The only difference this really makes is in coldboot attacks, but even then it's still not clear whether it makes the attacker's job slightly easier, or allows an attack that's otherwise not possible.
This logic is perfectly aligned with the Chromium threat model. Once an attacker gains administrator access, it is game over by definition.
I doubt this is an Edge-specific issue. Microsoft has no interest in making their browser less secure than its upstream.
> Why aren‘t physically-local attacks in Chrome’s threat model?
> We consider these attacks outside Chrome's threat model, because there is no way for Chrome (or any application) to defend against a malicious user who has managed to log into your device as you, or who can run software with the privileges of your operating system user account. Such an attacker can modify executables and DLLs, change environment variables like PATH, change configuration files, read any data your user account owns, email it to themselves, and so on. Such an attacker has total control over your device, and nothing Chrome can do would provide a serious guarantee of defense. This problem is not special to Chrome — all applications must trust the physically-local user.
It absolutely ain't Edge-specific. Firefox (AFAICT) also keeps stored passwords in clear-text unless encrypted with a passphrase (which is not the default on desktop; on Android there's a fingerprint/PIN check to access them, but I don't know offhand if there's any encryption involved with that).
Really this is true of most credentials stored within applications; unless you're providing a decryption key on open (whether explicitly or on OS-level login using some keychain mechanism), the stored credentials are probably plaintext.
Or unless you need to reenter password/offer fingerprint after certain amount of time. Which, I think, should be the actual standard, and typically is with the apps like Bitwarden.
It's a very standard defense-in-depth technique to put secrets between guard pages and only make the secret page readable when needed. That way any inadvertent access, be it programming error or exploit, simply causes a segfault, unless it's raced with a valid access (in a multithreaded or shm context) or the exploit explicitly changed the permission bits. Most memory disclosure vulnerabilities don't allow you to do that.
That being said any single password, when used, passes through so many layers and components that it's likely impossible to even just wipe the contaminated memory locations. But that's fine, the password database is opened for most of the browser's lifetime, any given password actively being used is a rare event in comparison.
In recent years we've also had browser-exploitable vulnerabilities that allowed reading arbitrary memory as a regular user, but slowly or without full control over the locations. I think wiping credentials as soon as possible after use is a very sensible precaution, even if it's only a moat.
I wonder about those kinds of exploits that sit on a webpage, but what stops someone from injecting their payload on a sites login page? JS can grab the password in plaintext in such a scenario, at which point the password manager does not save you. Can we normalize Passkey more?
I think the point is that you can have arbitrary website read the browser’s memory so example.com can read the password for example.org and example.net.
It's surprisingly hard to do the compiler or cpu may see a write without a read and optimize it away. Windows has a SecureZeroMemory and a few other barrier primitives but not all languages reach to it
Agreed. I keep seeing "high priority" "vulns" that require so much system access to actually exploit that they become pointless. If an untrusted process can read your memory or run as an administrator you have already lost.
It honestly feels like more and more "security" people and businesses have less interest in actually securing systems and more in marketing themselves and their business hence the tendency to make every niche attack into a five alarm fire.
Many of them also have vested interests in furthering corporate authoritarianism, which is why letting users have full control over their devices is considered a security risk to them.
Security isn't black and white. If i leave a post-it note of my logins on my monitor, that's definitely less safe than in a unlocked drawer, and so on.
If I leave a post-it note of passwords on my monitor inside a vault to which only I have access, it’s not a big deal. That’s the point of the “airtight hatch” metaphor.
I think we've moved away from the secure perimeter thinking and towards defense in depth - if that list of passwords helps you get somewhere other than the vault, removing the post-it improves security. Vaults get infiltrated all the time - and often in partial ways like being able to see into the vault but not reach in.
Defence in depth matters, but an analysis here shows that the same mechanism used to breach the outer layers (getting administrative access) can be used to breach the next layer (more thoroughly prodding Edge or Chrome to give up passwords).
Right; but in the scenario of this Tweek, you've invited someone untrustworthy into the vault and are then freaking out because they can see the post-it note of passwords. It is inherently irrational.
This issue is inherently unfixable by ANY password manager, because the process model of the underlying OS isn't itself secure. No obfuscation will work, because the password manager itself needs to de-obfuscation it before use (and that memory too is dump-able).
All adding in-memory obfuscation does it make ignorant people feel better, while not moving the security needle even an inch.
I think we’re largely in agreement. I do think there’s some benefit in reducing the amount of time that a password is in cleartext in memory. But it’s pretty far down the list.
> This issue is inherently unfixable by ANY password manager, because the process model of the underlying OS isn't itself secure
Usually the confidential bits are hardware isolated away from the supervisor (host kernel/OS) in Enclaves/TEEs, Realms, Secure Elements, Security chips, etc.
True. But then your hardware dies, and you're locked out of every account you own. It is objectively good security, but has a ton of usability headaches yet to be really solved.
I've seen orgs move to passkeys only, then offer reset-questions (e.g. city of first job, etc); because the Customer Service volume/workflow wasn't figured out.
oh lawd, yes it does come down to 'who has the power to reset your account', and very few people want to take the path of 'no one has the power' in the case of lost credentials.
I swear, people who idolize passkey security must never travel anywhere.
PS: "just have more devices with passkeys", they invariably say.
Yeah right because people are made of money, everyone has the forethought, and a 2nd laptop in the US is a great asset when you're in Poland and can't login anywhere.
I've been avoiding passkeys but more and more websites are trying to push them, and one website I use now requires them. I've already got a password manager! I don't need to change everything again!
The subject here is literally websites trying to push passkeys on users. That is who is asking us to.
About every week now Amazon tries to trick me into creating a passkey. It doesn't even ask, it just goes ahead and triggers my browser passkey creation mechanism without my consent. PayPal recently tried to force me to create one too and I had to kill and restart the app because that was the only way to skip it. I'll stick to my password with 2FA, thanks.
It's wildly obnoxious that browsers don't let you generally suppress these prompts.
And if you take the nuclear option and strip your browser of WebAuthn support, then you obviously can't use any passkeys, which doesn't work for me - I have two sites where I do want to use passkeys (because it's the only way to avoid SMS-based MFA on every login), but I never want to see passkey prompts for any other sites.
For people who only use passwords having an extra device can help too. Google does not necessarily permit a login with a backup code, so to me it seems ideal to grab a spare phone, log into important accounts, and store it with a trusted party/friend.
It could be very difficult to login to an account like Gmail from overseas in the event of PC+phone[+hardware key] theft. Maybe no big deal if you can port your number to a new phone right away. Or maybe the trusted friend can help (unless Google still finds the login suspicious after all, no idea there)
I travel a lot. By train, plane, and car. I also use passkeys when possible. I have multiple Yubikeys, stored in different locations. I also have a password manager, where I typically keep track of which logins aren’t yet backed up across physical tokens.
It takes a bit of effort, but it’s not impossible.
Yes, it means that in the event of catastrophic failure I might not be able to log in to some services until I get to one of the backups. I haven’t been able to imagine a scenario where that would be truly problematic.
>It is objectively good security, but has a ton of usability headaches yet to be really solved.
Thank you, then this is still true today?
Disappointing the rollout was botched (recall cross platform and password manager difficulties). Haven’t done research since but even with some new UIs and flows promoting passkeys in the past couple months, haven’t regained my trust either.
> If i leave a post-it note of my logins on my monitor, that's definitely less safe than in a unlocked drawer, and so on.
Having passwords on post-it notes does make certain types of attacks much easier. For instance, coworkers hacking other coworkers, or people burglarizing the office. None of which really apply to the "If an attacker gains administrative access on a terminal server" scenario.
Continuing the analogy, what Edge is doing is like leaving cash in unlocked cabinets inside a vault, and what Chrome's doing is locking those cabinets with a padlock. Sure, having the padlocks makes the cash more secure, but if someone went through all the effort into breaking the vault (terminal server), a padlock probably isn't going to stop them. This is especially true nowadays with AI coding agents and ready-made stealers available for sale online.
The way to think about security is as a system of layers, each of which filters out ever more sophisticated attackers.
We should care about all kinds of attackers, and not assume that the protections against the most sophisticated will obviate the protections against the least sophisticated.
The Swiss cheese model is what people use to sell you more 'security' related software systems that inherently involve more problems. (Also cheese is not very durable, even the kind without holes.)
isn't it at risk of any code pathway that somehow allows you exceed a buffer and read memory unbounded? Then a nefarious web page could capture that? That's a huge exposure surface.
All true, but it is still bad style. There is no need to keep decrypted passwords in memory the user hasn’t even used in the session (or after they logged in to a certain website).
So as a user, when I go to paste any password, I first need to log in with biometrics, yet any random user process can just snag the password from memory?
Reading arbitrary process memory can be done as a standard user. No admin needed. Any Win32 program can do it. You just can't access the memory from processes that are admin-level.
This is not true. The canonical way to prevent access is via PAGE_NOACCESS[1]. Obviously, running as admin or in kernel mode breaks the whole thing since you can re-call `VirtualProtect` on that page and open it up.
This is accurate as far as page protection goes. The problem is the largest threat model.
If Process A and Process B are running in the same user context on a desktop OS, PAGE_NOACCESS is not a strong boundary by itself. Process B may be able to obtain PROCESS_VM_OPERATION/PROCESS_VM_READ, change the page protection with VirtualProtectEx, inject code that calls VirtualProtect inside Process A, load a DLL, attach as a debugger, duplicate useful handles, or tamper with the executable. That's the problem with same-user process isolation, it is a hugely leaky abstraction. There is no magical "just set this bit" fix.
On a desktop OS, once an evil process runs under the same user context, you are relying on process DACLs, integrity levels, code-signing, anti-injection hardening, and file-system protections. You can plug one path and still have several others.
This comment feels like it's written by AI. Anyway, PAGE_GUARD helps you get around VirtualProtectEx, which is a very common way of detecting userspace cheats.
I'm not the other commenter (and I believe you that it's not AI), but I'd guess it's mostly the first line: a short affirmation followed by "The problem is ...." feels like the sort of formula the LLMs love to use. (Not trying to imply that there's anything inherently wrong with it, of course.)
While we're at it, I'm under the impression that the recent LLMs have also co-opted "genuinely", which I'll never forgive them for—first they stole my em-dashes, and now they're stealing my adverbs too?!
I do see how your comment is similar to AI writing (a couple other comments explain) but it did NOT set off my AI trigger. I think it's just clear writing.
Without context, sentences like this mean nothing. So it's borderline a non sequitur. A threat model can be literally anything. Me giving my PC to someone at Best Buy, letting my grandma write assembly, or throwing my PC out the window can be a "large threat model." Nonsense sentence.
> If Process A and Process B are running in the same user context on a desktop OS, PAGE_NOACCESS is not a strong boundary by itself. Process B may be able to obtain PROCESS_VM_OPERATION/PROCESS_VM_READ, change the page protection with VirtualProtectEx, inject code that calls VirtualProtect inside Process A, load a DLL, attach as a debugger, duplicate useful handles, or tamper with the executable.
To the uninitiated this seems right, but really there's so much glossing over, it feels written by a non-expert that just read the first chapter of a "hacking for dummies" book. I've written anti-cheats and have even done some some hardware stuff, so I say this with some degree of experience: writing a userspace hack/cheat is pretty hard without a zero-day. Most stuff won't easily get PROCESS_VM_OPERATION permissions, also those are (afaik) logged by the kernel, so you can easily see if some weird "DefinitelyNotACheat.exe" executable or "NotABadLibrary.dll" requested them, so it's a pretty janky way of getting access to memory you shouldn't.
> That's the problem with same-user process isolation, it is a hugely leaky abstraction. There is no magical "just set this bit" fix.
Again, this is a non sequitur. No one said (or at least I didn't) that there's a "magical" bit. You're not even arguing against a strawman, it's almost like we're having two different conversations.
> On a desktop OS, once an evil process runs under the same user context, you are relying on process DACLs, integrity levels, code-signing, anti-injection hardening, and file-system protections. You can plug one path and still have several others.
Also seems right, and it kinda' is, but code signing is notoriously easy to circumvent, "anti-injection hardening" can mean like three million different things, etc. I dunno, just sounds like someone that's never done this stuff before. Like, not bringing up Detours[1] when talking about "anti-injection" just seems like weirdly avoiding the ONE canonical way of doing this, which just about every single hacking/cracking book covers. Idk, weird omission.
Also, no one in their right mind would attach a debugger, as that's trivial to detect[2]. I guess it could be a decent proof of concept, but no serious hacker would ever go that route. (Also, if I remember correctly, you also need to ship some special DLLs that have the actual debugging helpers—and same with Detours, so might as well do that).
Just wanted to give my justification for the accusation. Maybe I'm wrong and maybe that's why I'm getting the downvotes, so my bad.
They also act as access alarms[1]. Why even comment if you didn't bother to read the docs?
> The PAGE_GUARD protection modifier establishes guard pages. Guard pages act as one-shot access alarms. For more information, see Creating Guard Pages.
Thankfully our recent experiences with OpenClaw have given us all a lot of faith that users are extremely diligent in what processes they allow access to what information.
This is a fallacious belief. While there is not point in obscurity, there is much value in not making it trivially easy to read passwords, as most exploits (especially of chromium) are not full user compromise, but the ability to massage some memory structures and read/write specific interesting bytes.
Additionally, the passwords could be kept encrypted in another process, and decrypted on demand, essentially a password vault. This lets you use techniques like biometric or physical button approval for password use, and reduces the likelihood of a browser memory dump containing passwords.
File audit capabilities in the OS can also be tuned so that only the vault application should be reading the vault file. Make info stealers job difficult.
They want obscurity and think it's security. Everything needed to get the passwords must be present in memory but they don't want to be able to actually see the passwords directly.
There's little hope of protecting against a snooper seeing the passwords you actually use, since they have to exist in plaintext at some point. But there's no reason to expose the entire password database when no passwords are even being used.
This is 100% that case. Basically every form (like this very one I'm typing in) is held in userspace memory un-encrypted. And yet lawyers and doctors and CIA operatives all use forms to type very sensitive stuff in.
It would be stupid, wasteful, and overly-complex to encrypt forms just in case some malicious process somehow got ring0 access. In that case, a keylogger is likely more useful anyway. And you're fucked even if you are encrypting stuff (as keys are likely also somewhere in memory[1] and they need to be—gasp—unencrypted). There's no free lunch.
Stupid Twitter thread meant to rage-bait for engagement.
[1] They could also be on disk or on some peripheral, but still fully readable by a motivated-enough hacker.
Sorry for off topic, is the situation the same with Safari on MacOS? I have to touch-id every time I fill the password, so it seems like it’s not available in-memory.
For reference, this is how Google says Chrome stores passwords encrypted in memory and uses an elevated service to prevent other processes from impersonating Chrome and gaining access to the plain text passwords: https://security.googleblog.com/2024/07/improving-security-o...
Since it's not been clearly stated: One attack vector might be that I step out to the bathroom for 5 minutes without locking computer, and evil hacker just dumps all my passwords before I come back.
I think it's worthwhile considering this. There's a reason why password managers ask for a master password or passkey after 10 minutes. Since I thought Chrome relied on an encrypted enclave, it isn't quite feasible to extract passwords easily even with root access.
Yes, you shouldn't leave your computer unattended. But that doesn't mean designing products that make exploiting the inevitable slipup fatal.
Password managers often go through quite some hassle to keep passwords 'safe' in memory. However, I often do not get the attack model of many of those tools. Tools like keepass e.g. go through quite to register a browser plugin. But then anyone with normal user rights can extract that key from the browser and do everything with it. Also this whole 'trust this browser' stuff of web apps seems strange if one e.g. can read the cookie store easily...
Cookies, if done correctly, will store a string that the server offered after a successful authentication - that string should have nothing to do with the password (it might contain some user information for logging/cross site tracking) but nothing sensitive.
With said cookie you can absolutely impersonate a user for while (potentially needing to evade user agent string checks and the like but often not)... but it will expire and then your access should be ended. If the site is well designed actions like password changing should also re-require the user's password instead of allowing anyone with just the cookie from proceeding with the action.
If it is done right cookies are pretty decently secure at keeping your secrets safe but, for convenience they do lower the security that could be accomplished with more involved techniques.
As an aside Oauth's key -> token approach is basically identical to password -> cookie (assuming best practices are in place).
There are (illegal) marketplaces initial access brokers sell session cookies on. Some companies try to defend against that by e.g. checking whether it's even possible that you travelled from place A to place B within a certain timeframe and, based on that, might invalidate your cookie. But then again attackers, depending on their sophistication, find their ways around it by ensuring they proxy their traffic via geographically close residential proxies, use the same OS and browser versions, etc.
The headline here says "stores in memory", which sounds pretty much identical to me. Can you elaborate on what you consider the difference between "loading" and "storing" into memory?
When someone says passwords are ‘stored’, the assumption will always be ‘stored on disk’. ‘stores in memory’ is not an accurate representation because memory is inherently volatile and they are loaded there temporarily. Plaintext on disk is egregious, plaintext in memory is considerably less so.
> When someone says passwords are ‘stored’, the assumption will always be ‘stored on disk’. ‘stores in memory’ is not an accurate representation
I mean, sure, if you literally ignore the words "in memory", but by that logic you could argue that "Microsoft Edge stores" is misleading because it sounds like it's talking about retail establishments that sell the web browser, which is equally nonsense. I don't find it plausible that you think most people would see "stores in memory" would mean "stores on disk" unless you think that they don't understand the difference between memory and disk, at which point I don't think that they would be here to misread the headline.
especially when the point of a password manager is to stick a plaintext string into a webpage, which then transmits the plain text to a remote server. passwords are just not a very good solution to keeping secrets.
To fair though, there are very few situations where the network is completely trustworthy, like your home network with no one else on it or a VPN direct to an HTTP server.
My understanding was that if you have a valid https session, you are good.
A really really untrustworthy network could MITM your SSL connections and impose itself in front of all of them (Cisco IronPort?) but I think even then your browser will complain unless you've installed a proxy that allows it or a custom root certificate.
If you're on prem or able to manipulate the machine into an OS of your choosing, yes. But with purely remote access to a device the disk is pretty decently secured (even if Window's ACLs are nightmareishly convoluted).
Correct me if I am wrong but chrome is-at least was- keeping passwords as raw text in Windows too. I got friend's forgotten password from Chrome on 2021 version
I've always hated that argument. Yes, if someone as access to your local file system, you are already SOL, but if that machine is part of an org, they aren't necessarily SOL except for now those plain text passwords can potentially be used for easier lateral movement to hit other, more privileged accounts (if you had access/had them saved in that password manager). At minimum, those credentials can now be used to phish the rest of your organization.
Stopping the spread is just as important as protecting any individual machine.
Edge is built by a company not focusing on user data-protection, so no surprise here. At least Brave and Firefox are usable and actual competitors, but have a business model based on user security rather than data.
mixed feelings on this, edge is supposed to store creds via DPAPI to the most part. you should also really not use password saving feature on edge (or any browser), it exposes you to a lot more threats that you need.
But.. saved passwords are not the same thing as "secrets" the browser uses. It has to be able to provide plain text passwords to websites. This is a really bad feature browsers should just not have to begin with, but they do, and I don't see a better way to use this.
In the past, they used to store the passwords in sqlite dbs, but now they've moved away from that at least.
From an attack perspective, there maybe some instances where you can dump memory, but you can't attach a debugger to the process without getting caught. so it does make a little bit of a difference there, but microsoft will probably tell you this isn't a security boundary that's being crossed. They can store it via DPAPI in lsass, and if lsass isolation is enabled (only on physical computers, default on win11) even SYSTEM privilege won't get you the credentials.
But what's the idea here, you have access to the browser, but you can't visit the site the password is saved for to make it "in use" and in plain text, so you can dump the password? I mean, even if you don't have access to the desktop, you can just start msedge.exe with the URL for the site as an argument and trigger the password retrieval.
Edge has done a lot to improve credential security, even DPAPI's existence itself is huge. If your research has meat, that's great but I don't see it here.
This feels like some "researcher" hyping themselves up to me, but I could be wrong.
Also, I really despise how they posted this on twitter, not even considering the political landmine there, I can't see the comments or threads on there without logging in. I can't visit the site on mobile without being redirected to download the app. I just wanted to mention that if you use X as a security professional in this day and age, my opinion of you drops by like 50% immediately. I don't care if you use bluesky, vk, telegram, discord,facebook, threads or whatever else, twitter is the worst place for you to share your work and you should know better.
In this day and time Microsoft should really know better. But I have seen this, and worse, happen over and over again in some fortune 500 companies with ERP and in-house systems.
I would think this is a local vulnerability assuming Windows works as other OSs.
For anyone that thinks this is an Edge-specific dunk, Chrome does not hash your passwords and they are cleartext in memory while Chrome is running (which for most users is always).
Password hashes are one-directional lossy storage. If a password manager "hashed your password" it would be essentially deleting your password and replacing it with something else which cannot be used to log into anything. The password MUST be recoverable to plain-text to replay it to a website.
But you're correct that Chrome, Firefox, Edge, Lastpass, BitWarden, even Keepass have the same issue. It is an Operating System limitation, not a password manager problem.
I never said that they should be hashed, just that they aren't. Just subtly pointing out what the tradeoffs are if you choose to use a password manager whose storage/access is basically always available.
At least with Keepass it's locked in an encrypted store and only available exactly when I need it to be. I can take other precautions if I want when I want to access it.
With your browser's password manager you're stuck with the slop you were given.
This is generally true of every application that handles sensitive data. Unless you explicitly clear that memory, it's likely to hang around forever.
For example, here is a 2019 writeup from KeePassXC with similar notes: https://keepassxc.org/blog/2019-02-21-memory-security/ - even though they explicitly clear sensitive data, there is still a window of opportunity.
During my time working on confidential computing, we had a variety of demos showing similar attacks against lots of different datastores, scripts, etc. That's just how computers work and your options are very limited if this is part of your threat model (imo just confidential computing and, if you can handle the performance hit, fully-homomorphic encryption).
Windows already has a secure kernel credential store, they could move the Edge password store there with a bit of effort, minimize the splash damage when you retrieve a single password to send over HTTP from the regular user space.
> Credential Guard prevents credential theft attacks by protecting NTLM password hashes, Kerberos Ticket Granting Tickets (TGTs), and credentials stored by applications as domain credentials.
> Credential Guard uses Virtualization-based security (VBS) to isolate secrets so that only privileged system software can access them.
Browser-based password management serves the purpose of locking users into a specific browser; I'd much rather have the freedom to switch browsers at will without the cognitive tax of securely moving all my creds every time I want to switch my main browser.
That's not what that is saying. It's saying don't use an _online_ password manager instead of the browser one. In the very opening they state that simple implementations are great and even lists some. Then the rest of the article dives specifically into online password managers, which are something else.
Out of curiosity, why KeePass versus Bitwarden? I've been using Bitwarden for years, but if there's a specific reason I should be using KeePass instead, I'm open to changing.
KeePass is just an encrypted database file with UI around it for usability. You can keep the db on a USB drive, sync it through a cloud storage, e-mail it to yourself, whatever ... It's really not that complicated. BitWarden is the above as a service, I reckon.
Nb. The above refers to KeePassX. No idea what the KeePass without the x is about.
Naming things. So hard.
No fancy browser plugins, the ability to autotype, the db file could be synced with anything you can sync files.
Working search - not sure about BW, but it's opensource implementation (Vaultwarden nowadays?) simply didn't allow to search for the fields you didn't scroll yet to.
The biggest problem is lack of multi-edit functionality - you need keep it in mind if you leave somehwere a copy running 24/7.
Bitwarden has taken investor money, sadly. It's still in good shape for the moment. But the time will come when they place profits above other needs; it's a matter of when, not if.
Luckily offering enterprise / credential sharing features is a decent freemium model. It still wins out in keeping compatibility with self hosted vaultwarden, are there other extensions that let you point to your own domain for the encrypted blob storage?
If it is a process, running in the same user context, with the ability to read/dump arbitrary memory -- As the KeePass database is decrypted it would "store all passwords in memory in plain text" too.
The fix isn't Edge Vs. Chrome. Vs KeePass Vs. Bitwarden, it is "How do I have my passwords exist in a different execution context than [evil process able to read all memory]?"
Android and iOS have an "answer" to this problem. Desktop OSs having all processes running side by side in the user's execution context, do not. It is only as secure as the least secure process running.
I was looking for an answer to this when it comes to using Edge password manager in particular, it uses Windows Hello as far as I know and while it does make 'synced' passkeys they don't seem to be usable anywhere than the original machine. Useful when reinstalling Windows at least.
This makes me miss running Qubes a few years ago, and keeping BitWarden in a separate VM from everything else. I've never felt as secure as when I had that setup.
Windows already has a secure kernel credential store, they could move the Edge password store there with a bit of effort, minimize the splash damage when you retrieve a single password to send over HTTP from the regular user space.
> Credential Guard prevents credential theft attacks by protecting NTLM password hashes, Kerberos Ticket Granting Tickets (TGTs), and credentials stored by applications as domain credentials.
> Credential Guard uses Virtualization-based security (VBS) to isolate secrets so that only privileged system software can access them.
I'm pretty sure macOS is more like iOS in this respect. At the very least, the passwords are typically secured biometrically and only the one being used is actually decrypted at the time of use.
I don't understand, who are all these people who care about security and at the same time are using Microsoft Edge. Could someone enlighten me? Does it have some specific features that somebody needs?
Linux stores plenty of passwords in clear text in /etc and $HOME and this is considered acceptable by most users. These same people also believe the TPM is a spy chip.
Really in /etc plain text? I could see some random app possibly doing that somewhere in ~/.config, but I don't think Linux itself stores passwords in plain text for systemwide use.
I think in general one should not assume anything in Edge is done correctly. Microsoft Edge is the place where things get tried out my Microsoft, that's why it changes so fast. It has a built-in updater that is not tied to Windows update, and as such they can iterate incredibly fast.
This feels like a case of "It rather involved being on the other side of this airtight hatchway"[1]. If you can read arbitrary process memory, you're probably also in a position to just dump out the passwords by pretending to be the user in question.
> If an attacker gains administrative access on a terminal server, they can access the memory of all logged‑on user processes.
If an attacker has administrative access, they can also attach a debugger to every chrome process and force it to decrypt all the passwords. The only difference this really makes is in coldboot attacks, but even then it's still not clear whether it makes the attacker's job slightly easier, or allows an attack that's otherwise not possible.
[1] https://devblogs.microsoft.com/oldnewthing/20060508-22/?p=31...
This logic is perfectly aligned with the Chromium threat model. Once an attacker gains administrator access, it is game over by definition.
I doubt this is an Edge-specific issue. Microsoft has no interest in making their browser less secure than its upstream.
> Why aren‘t physically-local attacks in Chrome’s threat model?
> We consider these attacks outside Chrome's threat model, because there is no way for Chrome (or any application) to defend against a malicious user who has managed to log into your device as you, or who can run software with the privileges of your operating system user account. Such an attacker can modify executables and DLLs, change environment variables like PATH, change configuration files, read any data your user account owns, email it to themselves, and so on. Such an attacker has total control over your device, and nothing Chrome can do would provide a serious guarantee of defense. This problem is not special to Chrome — all applications must trust the physically-local user.
https://chromium.googlesource.com/chromium/src/+/148.0.7778....
> I doubt this is an Edge-specific issue.
It absolutely ain't Edge-specific. Firefox (AFAICT) also keeps stored passwords in clear-text unless encrypted with a passphrase (which is not the default on desktop; on Android there's a fingerprint/PIN check to access them, but I don't know offhand if there's any encryption involved with that).
Really this is true of most credentials stored within applications; unless you're providing a decryption key on open (whether explicitly or on OS-level login using some keychain mechanism), the stored credentials are probably plaintext.
Or unless you need to reenter password/offer fingerprint after certain amount of time. Which, I think, should be the actual standard, and typically is with the apps like Bitwarden.
It's a very standard defense-in-depth technique to put secrets between guard pages and only make the secret page readable when needed. That way any inadvertent access, be it programming error or exploit, simply causes a segfault, unless it's raced with a valid access (in a multithreaded or shm context) or the exploit explicitly changed the permission bits. Most memory disclosure vulnerabilities don't allow you to do that.
That being said any single password, when used, passes through so many layers and components that it's likely impossible to even just wipe the contaminated memory locations. But that's fine, the password database is opened for most of the browser's lifetime, any given password actively being used is a rare event in comparison.
> It's a very standard defense-in-depth technique
Is there any software we’d be aware of which uses this technique
Wouldn’t a guard page be readable in Linux with /proc/self/mem ? (at least read only pages are writable with it)
Come on, they could still get a blood sample to really verify that its the user
In recent years we've also had browser-exploitable vulnerabilities that allowed reading arbitrary memory as a regular user, but slowly or without full control over the locations. I think wiping credentials as soon as possible after use is a very sensible precaution, even if it's only a moat.
I wonder about those kinds of exploits that sit on a webpage, but what stops someone from injecting their payload on a sites login page? JS can grab the password in plaintext in such a scenario, at which point the password manager does not save you. Can we normalize Passkey more?
I think the point is that you can have arbitrary website read the browser’s memory so example.com can read the password for example.org and example.net.
It's surprisingly hard to do the compiler or cpu may see a write without a read and optimize it away. Windows has a SecureZeroMemory and a few other barrier primitives but not all languages reach to it
Agreed. I keep seeing "high priority" "vulns" that require so much system access to actually exploit that they become pointless. If an untrusted process can read your memory or run as an administrator you have already lost.
It honestly feels like more and more "security" people and businesses have less interest in actually securing systems and more in marketing themselves and their business hence the tendency to make every niche attack into a five alarm fire.
Many of them also have vested interests in furthering corporate authoritarianism, which is why letting users have full control over their devices is considered a security risk to them.
Security isn't black and white. If i leave a post-it note of my logins on my monitor, that's definitely less safe than in a unlocked drawer, and so on.
If I leave a post-it note of passwords on my monitor inside a vault to which only I have access, it’s not a big deal. That’s the point of the “airtight hatch” metaphor.
I think we've moved away from the secure perimeter thinking and towards defense in depth - if that list of passwords helps you get somewhere other than the vault, removing the post-it improves security. Vaults get infiltrated all the time - and often in partial ways like being able to see into the vault but not reach in.
Defence in depth matters, but an analysis here shows that the same mechanism used to breach the outer layers (getting administrative access) can be used to breach the next layer (more thoroughly prodding Edge or Chrome to give up passwords).
Right; but in the scenario of this Tweek, you've invited someone untrustworthy into the vault and are then freaking out because they can see the post-it note of passwords. It is inherently irrational.
This issue is inherently unfixable by ANY password manager, because the process model of the underlying OS isn't itself secure. No obfuscation will work, because the password manager itself needs to de-obfuscation it before use (and that memory too is dump-able).
All adding in-memory obfuscation does it make ignorant people feel better, while not moving the security needle even an inch.
I think we’re largely in agreement. I do think there’s some benefit in reducing the amount of time that a password is in cleartext in memory. But it’s pretty far down the list.
> This issue is inherently unfixable by ANY password manager, because the process model of the underlying OS isn't itself secure
Usually the confidential bits are hardware isolated away from the supervisor (host kernel/OS) in Enclaves/TEEs, Realms, Secure Elements, Security chips, etc.
One more reason to use hardware-bound passkeys and not passwords.
True. But then your hardware dies, and you're locked out of every account you own. It is objectively good security, but has a ton of usability headaches yet to be really solved.
I've seen orgs move to passkeys only, then offer reset-questions (e.g. city of first job, etc); because the Customer Service volume/workflow wasn't figured out.
oh lawd, yes it does come down to 'who has the power to reset your account', and very few people want to take the path of 'no one has the power' in the case of lost credentials.
At my work we required a complex password <15 characters lower + cap, number and symbols.
Updated to Windows Hello and passkey.
Now I can use a 4 digit pin to login.
>your hardware dies
Or your backpack gets stolen.
Oops.
I swear, people who idolize passkey security must never travel anywhere.
PS: "just have more devices with passkeys", they invariably say.
Yeah right because people are made of money, everyone has the forethought, and a 2nd laptop in the US is a great asset when you're in Poland and can't login anywhere.
I've been avoiding passkeys but more and more websites are trying to push them, and one website I use now requires them. I've already got a password manager! I don't need to change everything again!
Your password manager almost certainly already has baked-in passkey support.
It does, but what's your point? Why should I redo everything?
Nobody is asking you to?
The subject here is literally websites trying to push passkeys on users. That is who is asking us to.
About every week now Amazon tries to trick me into creating a passkey. It doesn't even ask, it just goes ahead and triggers my browser passkey creation mechanism without my consent. PayPal recently tried to force me to create one too and I had to kill and restart the app because that was the only way to skip it. I'll stick to my password with 2FA, thanks.
It's wildly obnoxious that browsers don't let you generally suppress these prompts.
And if you take the nuclear option and strip your browser of WebAuthn support, then you obviously can't use any passkeys, which doesn't work for me - I have two sites where I do want to use passkeys (because it's the only way to avoid SMS-based MFA on every login), but I never want to see passkey prompts for any other sites.
>"just have more devices with passkeys"
Confirms that strategy then
For people who only use passwords having an extra device can help too. Google does not necessarily permit a login with a backup code, so to me it seems ideal to grab a spare phone, log into important accounts, and store it with a trusted party/friend.
It could be very difficult to login to an account like Gmail from overseas in the event of PC+phone[+hardware key] theft. Maybe no big deal if you can port your number to a new phone right away. Or maybe the trusted friend can help (unless Google still finds the login suspicious after all, no idea there)
I travel a lot. By train, plane, and car. I also use passkeys when possible. I have multiple Yubikeys, stored in different locations. I also have a password manager, where I typically keep track of which logins aren’t yet backed up across physical tokens.
It takes a bit of effort, but it’s not impossible.
Yes, it means that in the event of catastrophic failure I might not be able to log in to some services until I get to one of the backups. I haven’t been able to imagine a scenario where that would be truly problematic.
>It is objectively good security, but has a ton of usability headaches yet to be really solved.
Thank you, then this is still true today?
Disappointing the rollout was botched (recall cross platform and password manager difficulties). Haven’t done research since but even with some new UIs and flows promoting passkeys in the past couple months, haven’t regained my trust either.
> If i leave a post-it note of my logins on my monitor, that's definitely less safe than in a unlocked drawer, and so on.
Having passwords on post-it notes does make certain types of attacks much easier. For instance, coworkers hacking other coworkers, or people burglarizing the office. None of which really apply to the "If an attacker gains administrative access on a terminal server" scenario.
Continuing the analogy, what Edge is doing is like leaving cash in unlocked cabinets inside a vault, and what Chrome's doing is locking those cabinets with a padlock. Sure, having the padlocks makes the cash more secure, but if someone went through all the effort into breaking the vault (terminal server), a padlock probably isn't going to stop them. This is especially true nowadays with AI coding agents and ready-made stealers available for sale online.
> Having passwords on post-it notes does make certain types of attacks much easier.
It also makes other attacks much harder. Namely I don't need to worry about some zero-day in my password manager.
The way to think about security is as a system of layers, each of which filters out ever more sophisticated attackers.
We should care about all kinds of attackers, and not assume that the protections against the most sophisticated will obviate the protections against the least sophisticated.
The Swiss cheese model. Each single layer has holes, but when stacked the combined hole area is minimized https://en.wikipedia.org/wiki/Swiss_cheese_model
The Swiss cheese model is what people use to sell you more 'security' related software systems that inherently involve more problems. (Also cheese is not very durable, even the kind without holes.)
Swiss cheese applies to more than just security systems.
Hiking with two GPS-capable devices is Swiss cheese.
That was an enlightening read, considering the colloquial meaning of "your firewall security is like Swiss cheese"
https://en.wiktionary.org/wiki/Swiss_cheese#Noun
What's next? A system so secure that you can drive a truck through it? A honeypot in the center of a wasp nest?
Okay. Can you describe an attack / threat model where it would matter in this particular case?
isn't it at risk of any code pathway that somehow allows you exceed a buffer and read memory unbounded? Then a nefarious web page could capture that? That's a huge exposure surface.
I'm pretty sure a read exploit in a web page wouldn't be in the same process as the passwords.
If you can cross over to the main Edge process, you can probably get it to remove any encryption it applied itself.
https://en.wikipedia.org/wiki/Spectre_(security_vulnerabilit...
Have we already forgotten Cloudbleed [0]?
[0] https://en.wikipedia.org/wiki/Cloudbleed
All true, but it is still bad style. There is no need to keep decrypted passwords in memory the user hasn’t even used in the session (or after they logged in to a certain website).
So as a user, when I go to paste any password, I first need to log in with biometrics, yet any random user process can just snag the password from memory?
What am I missing here?
So you're saying it's an Edge case?
Reading arbitrary process memory can be done as a standard user. No admin needed. Any Win32 program can do it. You just can't access the memory from processes that are admin-level.
This is not true. The canonical way to prevent access is via PAGE_NOACCESS[1]. Obviously, running as admin or in kernel mode breaks the whole thing since you can re-call `VirtualProtect` on that page and open it up.
[1] https://learn.microsoft.com/en-us/windows/win32/memory/memor...
This is accurate as far as page protection goes. The problem is the largest threat model.
If Process A and Process B are running in the same user context on a desktop OS, PAGE_NOACCESS is not a strong boundary by itself. Process B may be able to obtain PROCESS_VM_OPERATION/PROCESS_VM_READ, change the page protection with VirtualProtectEx, inject code that calls VirtualProtect inside Process A, load a DLL, attach as a debugger, duplicate useful handles, or tamper with the executable. That's the problem with same-user process isolation, it is a hugely leaky abstraction. There is no magical "just set this bit" fix.
On a desktop OS, once an evil process runs under the same user context, you are relying on process DACLs, integrity levels, code-signing, anti-injection hardening, and file-system protections. You can plug one path and still have several others.
This comment feels like it's written by AI. Anyway, PAGE_GUARD helps you get around VirtualProtectEx, which is a very common way of detecting userspace cheats.
> This comment feels like it's written by AI.
Why exactly? I'm genuinely asking, because I feel like I get this a lot, and it is pretty frustrating.
I'm not the other commenter (and I believe you that it's not AI), but I'd guess it's mostly the first line: a short affirmation followed by "The problem is ...." feels like the sort of formula the LLMs love to use. (Not trying to imply that there's anything inherently wrong with it, of course.)
While we're at it, I'm under the impression that the recent LLMs have also co-opted "genuinely", which I'll never forgive them for—first they stole my em-dashes, and now they're stealing my adverbs too?!
Thanks for the explanation. Yeah, I use "genuinely" and "honestly" far too much; and often in odd places. It is a bad habit.
As to that comment's tone, my entire comment history is visible going back years. I'd invite people to peruse it.
I do see how your comment is similar to AI writing (a couple other comments explain) but it did NOT set off my AI trigger. I think it's just clear writing.
> The problem is the largest threat model.
Without context, sentences like this mean nothing. So it's borderline a non sequitur. A threat model can be literally anything. Me giving my PC to someone at Best Buy, letting my grandma write assembly, or throwing my PC out the window can be a "large threat model." Nonsense sentence.
> If Process A and Process B are running in the same user context on a desktop OS, PAGE_NOACCESS is not a strong boundary by itself. Process B may be able to obtain PROCESS_VM_OPERATION/PROCESS_VM_READ, change the page protection with VirtualProtectEx, inject code that calls VirtualProtect inside Process A, load a DLL, attach as a debugger, duplicate useful handles, or tamper with the executable.
To the uninitiated this seems right, but really there's so much glossing over, it feels written by a non-expert that just read the first chapter of a "hacking for dummies" book. I've written anti-cheats and have even done some some hardware stuff, so I say this with some degree of experience: writing a userspace hack/cheat is pretty hard without a zero-day. Most stuff won't easily get PROCESS_VM_OPERATION permissions, also those are (afaik) logged by the kernel, so you can easily see if some weird "DefinitelyNotACheat.exe" executable or "NotABadLibrary.dll" requested them, so it's a pretty janky way of getting access to memory you shouldn't.
> That's the problem with same-user process isolation, it is a hugely leaky abstraction. There is no magical "just set this bit" fix.
Again, this is a non sequitur. No one said (or at least I didn't) that there's a "magical" bit. You're not even arguing against a strawman, it's almost like we're having two different conversations.
> On a desktop OS, once an evil process runs under the same user context, you are relying on process DACLs, integrity levels, code-signing, anti-injection hardening, and file-system protections. You can plug one path and still have several others.
Also seems right, and it kinda' is, but code signing is notoriously easy to circumvent, "anti-injection hardening" can mean like three million different things, etc. I dunno, just sounds like someone that's never done this stuff before. Like, not bringing up Detours[1] when talking about "anti-injection" just seems like weirdly avoiding the ONE canonical way of doing this, which just about every single hacking/cracking book covers. Idk, weird omission.
Also, no one in their right mind would attach a debugger, as that's trivial to detect[2]. I guess it could be a decent proof of concept, but no serious hacker would ever go that route. (Also, if I remember correctly, you also need to ship some special DLLs that have the actual debugging helpers—and same with Detours, so might as well do that).
Just wanted to give my justification for the accusation. Maybe I'm wrong and maybe that's why I'm getting the downvotes, so my bad.
[1] https://github.com/microsoft/detours
[2] https://learn.microsoft.com/en-us/windows/win32/api/debugapi...
Guard pages are one-shot exceptions used for growing the stack.
They also act as access alarms[1]. Why even comment if you didn't bother to read the docs?
> The PAGE_GUARD protection modifier establishes guard pages. Guard pages act as one-shot access alarms. For more information, see Creating Guard Pages.
[1] https://learn.microsoft.com/en-us/windows/win32/api/memoryap...
And if the malware is running as admin, you’re pretty fucked either way
Thankfully our recent experiences with OpenClaw have given us all a lot of faith that users are extremely diligent in what processes they allow access to what information.
This is a fallacious belief. While there is not point in obscurity, there is much value in not making it trivially easy to read passwords, as most exploits (especially of chromium) are not full user compromise, but the ability to massage some memory structures and read/write specific interesting bytes.
Additionally, the passwords could be kept encrypted in another process, and decrypted on demand, essentially a password vault. This lets you use techniques like biometric or physical button approval for password use, and reduces the likelihood of a browser memory dump containing passwords.
File audit capabilities in the OS can also be tuned so that only the vault application should be reading the vault file. Make info stealers job difficult.
>If you can read arbitrary process memory, you're probably also in a position to just dump out the passwords by pretending to be the user in question.
This is the load bearing argument and it is false.
There are plenty of circumstances were you can grab a piece of process memory but not all of it.
There are plenty more circumstances where you can grab process memory but not kernel memory.
There are plenty more (almost all) where you can dump kernel and process memory but you can't access the keys stored in the TPM module.
Leaving the door open for anyone with the smallest exploit is stupid and bad security.
They want obscurity and think it's security. Everything needed to get the passwords must be present in memory but they don't want to be able to actually see the passwords directly.
There's little hope of protecting against a snooper seeing the passwords you actually use, since they have to exist in plaintext at some point. But there's no reason to expose the entire password database when no passwords are even being used.
This is 100% that case. Basically every form (like this very one I'm typing in) is held in userspace memory un-encrypted. And yet lawyers and doctors and CIA operatives all use forms to type very sensitive stuff in.
It would be stupid, wasteful, and overly-complex to encrypt forms just in case some malicious process somehow got ring0 access. In that case, a keylogger is likely more useful anyway. And you're fucked even if you are encrypting stuff (as keys are likely also somewhere in memory[1] and they need to be—gasp—unencrypted). There's no free lunch.
Stupid Twitter thread meant to rage-bait for engagement.
[1] They could also be on disk or on some peripheral, but still fully readable by a motivated-enough hacker.
Sorry for off topic, is the situation the same with Safari on MacOS? I have to touch-id every time I fill the password, so it seems like it’s not available in-memory.
For reference, this is how Google says Chrome stores passwords encrypted in memory and uses an elevated service to prevent other processes from impersonating Chrome and gaining access to the plain text passwords: https://security.googleblog.com/2024/07/improving-security-o...
That appears to be storage at rest (on disk), rather than in memory.
Since it's not been clearly stated: One attack vector might be that I step out to the bathroom for 5 minutes without locking computer, and evil hacker just dumps all my passwords before I come back.
I think it's worthwhile considering this. There's a reason why password managers ask for a master password or passkey after 10 minutes. Since I thought Chrome relied on an encrypted enclave, it isn't quite feasible to extract passwords easily even with root access.
Yes, you shouldn't leave your computer unattended. But that doesn't mean designing products that make exploiting the inevitable slipup fatal.
Does this tool access an Edge instance running on the same machine? Couldn't you then just simply export all saved passwords anyway?
https://support.microsoft.com/en-us/topic/export-passwords-i...
Password managers often go through quite some hassle to keep passwords 'safe' in memory. However, I often do not get the attack model of many of those tools. Tools like keepass e.g. go through quite to register a browser plugin. But then anyone with normal user rights can extract that key from the browser and do everything with it. Also this whole 'trust this browser' stuff of web apps seems strange if one e.g. can read the cookie store easily...
Cookies, if done correctly, will store a string that the server offered after a successful authentication - that string should have nothing to do with the password (it might contain some user information for logging/cross site tracking) but nothing sensitive.
With said cookie you can absolutely impersonate a user for while (potentially needing to evade user agent string checks and the like but often not)... but it will expire and then your access should be ended. If the site is well designed actions like password changing should also re-require the user's password instead of allowing anyone with just the cookie from proceeding with the action.
If it is done right cookies are pretty decently secure at keeping your secrets safe but, for convenience they do lower the security that could be accomplished with more involved techniques.
As an aside Oauth's key -> token approach is basically identical to password -> cookie (assuming best practices are in place).
There are (illegal) marketplaces initial access brokers sell session cookies on. Some companies try to defend against that by e.g. checking whether it's even possible that you travelled from place A to place B within a certain timeframe and, based on that, might invalidate your cookie. But then again attackers, depending on their sophistication, find their ways around it by ensuring they proxy their traffic via geographically close residential proxies, use the same OS and browser versions, etc.
Google now wants to bind credentials to a device by storing the secret in the TPM: https://blog.google/security/protecting-cookies-with-device-...
https://xcancel.com/L1v1ng0ffTh3L4N/status/20513083298807197...
Anyone have a link to the source code for this .exe? Would love to see _how_ it's extracting them.
Yeah, you can probably do the same thing to pam on linux... just attach gdb to openssh or your getty login process.
To be fair, 'loads into memory' and 'stores' are not the same thing.
The headline here says "stores in memory", which sounds pretty much identical to me. Can you elaborate on what you consider the difference between "loading" and "storing" into memory?
When someone says passwords are ‘stored’, the assumption will always be ‘stored on disk’. ‘stores in memory’ is not an accurate representation because memory is inherently volatile and they are loaded there temporarily. Plaintext on disk is egregious, plaintext in memory is considerably less so.
> When someone says passwords are ‘stored’, the assumption will always be ‘stored on disk’. ‘stores in memory’ is not an accurate representation
I mean, sure, if you literally ignore the words "in memory", but by that logic you could argue that "Microsoft Edge stores" is misleading because it sounds like it's talking about retail establishments that sell the web browser, which is equally nonsense. I don't find it plausible that you think most people would see "stores in memory" would mean "stores on disk" unless you think that they don't understand the difference between memory and disk, at which point I don't think that they would be here to misread the headline.
especially when the point of a password manager is to stick a plaintext string into a webpage, which then transmits the plain text to a remote server. passwords are just not a very good solution to keeping secrets.
Never enter your password into a website that doesn't use https.
*over any untrustworthy network.
To fair though, there are very few situations where the network is completely trustworthy, like your home network with no one else on it or a VPN direct to an HTTP server.
My understanding was that if you have a valid https session, you are good.
A really really untrustworthy network could MITM your SSL connections and impose itself in front of all of them (Cisco IronPort?) but I think even then your browser will complain unless you've installed a proxy that allows it or a custom root certificate.
That's kinda stupid. The passwords could get swapped to disk in the swap file in plaintext when memory is low by the OS.
You say this as if accessing that file was any easier than accessing memory.
If I have a disk image or access to the physical drive, it's trivial. This means they can no longer be considered encrypted at rest.
If you're on prem or able to manipulate the machine into an OS of your choosing, yes. But with purely remote access to a device the disk is pretty decently secured (even if Window's ACLs are nightmareishly convoluted).
If your computer storage is not fully encrypted you have bigger worries than the swap file.
It is when the computer is off.
Correct me if I am wrong but chrome is-at least was- keeping passwords as raw text in Windows too. I got friend's forgotten password from Chrome on 2021 version
Yeah it's been years but I remember seeing arguments with Google devs saying if someone had access to your local file system, you're already SOL.
I've always hated that argument. Yes, if someone as access to your local file system, you are already SOL, but if that machine is part of an org, they aren't necessarily SOL except for now those plain text passwords can potentially be used for easier lateral movement to hit other, more privileged accounts (if you had access/had them saved in that password manager). At minimum, those credentials can now be used to phish the rest of your organization.
Stopping the spread is just as important as protecting any individual machine.
Chrome added app bound encryption of cookie files in 2024.
The only important question is: does Chrome store passwords in the same way as Edge?
The real mistake is that we are still using simple password authentication instead of challenge-response or public key authentication.
Edge is built by a company not focusing on user data-protection, so no surprise here. At least Brave and Firefox are usable and actual competitors, but have a business model based on user security rather than data.
mixed feelings on this, edge is supposed to store creds via DPAPI to the most part. you should also really not use password saving feature on edge (or any browser), it exposes you to a lot more threats that you need.
But.. saved passwords are not the same thing as "secrets" the browser uses. It has to be able to provide plain text passwords to websites. This is a really bad feature browsers should just not have to begin with, but they do, and I don't see a better way to use this.
In the past, they used to store the passwords in sqlite dbs, but now they've moved away from that at least.
From an attack perspective, there maybe some instances where you can dump memory, but you can't attach a debugger to the process without getting caught. so it does make a little bit of a difference there, but microsoft will probably tell you this isn't a security boundary that's being crossed. They can store it via DPAPI in lsass, and if lsass isolation is enabled (only on physical computers, default on win11) even SYSTEM privilege won't get you the credentials.
But what's the idea here, you have access to the browser, but you can't visit the site the password is saved for to make it "in use" and in plain text, so you can dump the password? I mean, even if you don't have access to the desktop, you can just start msedge.exe with the URL for the site as an argument and trigger the password retrieval.
Edge has done a lot to improve credential security, even DPAPI's existence itself is huge. If your research has meat, that's great but I don't see it here.
This feels like some "researcher" hyping themselves up to me, but I could be wrong.
Also, I really despise how they posted this on twitter, not even considering the political landmine there, I can't see the comments or threads on there without logging in. I can't visit the site on mobile without being redirected to download the app. I just wanted to mention that if you use X as a security professional in this day and age, my opinion of you drops by like 50% immediately. I don't care if you use bluesky, vk, telegram, discord,facebook, threads or whatever else, twitter is the worst place for you to share your work and you should know better.
You are absolutely right, having copilot does not help at all here.
And firefox stores them unencrypted by default
A reminder that Edge is just Chromium plus some Microsoft hooks for automated SSO.
In this day and time Microsoft should really know better. But I have seen this, and worse, happen over and over again in some fortune 500 companies with ERP and in-house systems.
I would think this is a local vulnerability assuming Windows works as other OSs.
For anyone that thinks this is an Edge-specific dunk, Chrome does not hash your passwords and they are cleartext in memory while Chrome is running (which for most users is always).
Password hashes are one-directional lossy storage. If a password manager "hashed your password" it would be essentially deleting your password and replacing it with something else which cannot be used to log into anything. The password MUST be recoverable to plain-text to replay it to a website.
But you're correct that Chrome, Firefox, Edge, Lastpass, BitWarden, even Keepass have the same issue. It is an Operating System limitation, not a password manager problem.
I think the catch is whether the passwords are unencrypted in memory constantly, or only during a short period when the password is being used?
I never said that they should be hashed, just that they aren't. Just subtly pointing out what the tradeoffs are if you choose to use a password manager whose storage/access is basically always available.
At least with Keepass it's locked in an encrypted store and only available exactly when I need it to be. I can take other precautions if I want when I want to access it.
With your browser's password manager you're stuck with the slop you were given.
This is generally true of every application that handles sensitive data. Unless you explicitly clear that memory, it's likely to hang around forever.
For example, here is a 2019 writeup from KeePassXC with similar notes: https://keepassxc.org/blog/2019-02-21-memory-security/ - even though they explicitly clear sensitive data, there is still a window of opportunity.
During my time working on confidential computing, we had a variety of demos showing similar attacks against lots of different datastores, scripts, etc. That's just how computers work and your options are very limited if this is part of your threat model (imo just confidential computing and, if you can handle the performance hit, fully-homomorphic encryption).
Windows already has a secure kernel credential store, they could move the Edge password store there with a bit of effort, minimize the splash damage when you retrieve a single password to send over HTTP from the regular user space.
> Credential Guard prevents credential theft attacks by protecting NTLM password hashes, Kerberos Ticket Granting Tickets (TGTs), and credentials stored by applications as domain credentials.
> Credential Guard uses Virtualization-based security (VBS) to isolate secrets so that only privileged system software can access them.
https://learn.microsoft.com/en-us/windows/security/identity-...
Please use a dedicated password manager, instead of a browser-based one. KeePass is likely the best going forward.
@taviso had claimed the exact opposite: https://lock.cmpxchg8b.com/passmgrs.html
EDIT: Yes, he claimed that for online password managers, not keepass. I thought the argument was about password managers in general.
Where?
> Good examples of simple and safe password managers are keepass and keepassx
Browser-based password management serves the purpose of locking users into a specific browser; I'd much rather have the freedom to switch browsers at will without the cognitive tax of securely moving all my creds every time I want to switch my main browser.
I agree. It's especially problematic when you use different browsers on different devices and operating systems.
That's not what that is saying. It's saying don't use an _online_ password manager instead of the browser one. In the very opening they state that simple implementations are great and even lists some. Then the rest of the article dives specifically into online password managers, which are something else.
You're right. Edited my comment.
Out of curiosity, why KeePass versus Bitwarden? I've been using Bitwarden for years, but if there's a specific reason I should be using KeePass instead, I'm open to changing.
KeePass is just an encrypted database file with UI around it for usability. You can keep the db on a USB drive, sync it through a cloud storage, e-mail it to yourself, whatever ... It's really not that complicated. BitWarden is the above as a service, I reckon.
Nb. The above refers to KeePassX. No idea what the KeePass without the x is about. Naming things. So hard.
Bitwarden is cloud bases keepass is local
It's a program with a file database.
No fancy browser plugins, the ability to autotype, the db file could be synced with anything you can sync files.
Working search - not sure about BW, but it's opensource implementation (Vaultwarden nowadays?) simply didn't allow to search for the fields you didn't scroll yet to.
The biggest problem is lack of multi-edit functionality - you need keep it in mind if you leave somehwere a copy running 24/7.
Bitwarden has taken investor money, sadly. It's still in good shape for the moment. But the time will come when they place profits above other needs; it's a matter of when, not if.
Luckily offering enterprise / credential sharing features is a decent freemium model. It still wins out in keeping compatibility with self hosted vaultwarden, are there other extensions that let you point to your own domain for the encrypted blob storage?
If it is a process, running in the same user context, with the ability to read/dump arbitrary memory -- As the KeePass database is decrypted it would "store all passwords in memory in plain text" too.
The fix isn't Edge Vs. Chrome. Vs KeePass Vs. Bitwarden, it is "How do I have my passwords exist in a different execution context than [evil process able to read all memory]?"
Android and iOS have an "answer" to this problem. Desktop OSs having all processes running side by side in the user's execution context, do not. It is only as secure as the least secure process running.
Windows 11* and MacOS also do the job as long as you're using hardware bound passkeys.
* I don't want to speak past my own experience so checking my work, Windows can store passkeys in a TPM if available but falls back to storing on disk... https://helgeklein.com/blog/checking-windows-hello-for-busin...
I was looking for an answer to this when it comes to using Edge password manager in particular, it uses Windows Hello as far as I know and while it does make 'synced' passkeys they don't seem to be usable anywhere than the original machine. Useful when reinstalling Windows at least.
https://yourpasskeyisweak.com does not mention Edge.
This makes me miss running Qubes a few years ago, and keeping BitWarden in a separate VM from everything else. I've never felt as secure as when I had that setup.
Windows already has a secure kernel credential store, they could move the Edge password store there with a bit of effort, minimize the splash damage when you retrieve a single password to send over HTTP from the regular user space.
> Credential Guard prevents credential theft attacks by protecting NTLM password hashes, Kerberos Ticket Granting Tickets (TGTs), and credentials stored by applications as domain credentials.
> Credential Guard uses Virtualization-based security (VBS) to isolate secrets so that only privileged system software can access them.
https://learn.microsoft.com/en-us/windows/security/identity-...
I'm pretty sure macOS is more like iOS in this respect. At the very least, the passwords are typically secured biometrically and only the one being used is actually decrypted at the time of use.
I don't understand, who are all these people who care about security and at the same time are using Microsoft Edge. Could someone enlighten me? Does it have some specific features that somebody needs?
Its Microsoft doing Microsoft things
Linux stores plenty of passwords in clear text in /etc and $HOME and this is considered acceptable by most users. These same people also believe the TPM is a spy chip.
You really need to upgrade to UNIX Version 6 or later. Some of the improvements since 1974 are well worth having.
Really in /etc plain text? I could see some random app possibly doing that somewhere in ~/.config, but I don't think Linux itself stores passwords in plain text for systemwide use.
I think the commenter means that some Linux applications store the passwords they need for access to external resources in plain text.
You know `/etc/passwd` doesn't really have passwords in it.
I thought Linux stored plain text credentials owned by root that require elevated permissions.
> Linux stores plenty of passwords in plain text in /etc
That's gonna be a big ol' [CITATION NEEDED] from me, dawg.
Wifi passwords in /etc/netplan files, is one I can think of.
I haven't solved the problem of sensitive .env files sitting around on my computer.
`sops exec-env`
I have an alias set for when I'm working with opentofu:
`alias tfenter='sops exec-env secrets.yaml "/bin/bash"'`
I encrypt with openbao's transit engine and backup age key kept in a password manager, so no secrets live on disk.
My brain stores all my passwords in memory in clear text too
I think in general one should not assume anything in Edge is done correctly. Microsoft Edge is the place where things get tried out my Microsoft, that's why it changes so fast. It has a built-in updater that is not tied to Windows update, and as such they can iterate incredibly fast.
Why wouldn't it? What else would you expect from the p̶e̶o̶p̶l̶e̶ masochists who subjected us to internet explorer