If one chats/mails with a person using Windows, despite using secure private protocols, every message will be stored by Microsoft’s Windoze Recall. Either I’m missing something but this feature seems like the most grotesque breach in online privacy/security.
What are ways to avoid this except for using obfuscated text?
You can’t, at that point you assume your correspondent is compromised. It’s not just recall but also malware and credential stealers. Doesn’t matter if recall is taking screenshots, if the messaging client itself is pwned via malware then they have full access to as much history as is available.
This is a perfect example of why all of these privacy intrusion practices should be illegal. The same goes for services like Gmail. I use my own email server because I don’t want Google reading my messages. But it doesn’t matter, because everyone else uses Gmail, so any time I communicate with someone, Google reads my emails, despite the fact that I never agreed to their oppressive ToS. It’s a blatant violation of our right to privacy.
But it doesn’t matter, because everyone else uses Gmail, so any time I communicate with someone, Google reads my emails, despite the fact that I never agreed to their oppressive ToS.
That’s avoidable by PGP encrypting your emails though. But I’m sure you know that, and I’m sure you meant that getting most people to use PGP is a pipe dream.
Even if you got them using PGP somehow, there’s always a risk. Apps designed to upload screenshots, share contacts or simple human errors like “hey did you hear X saying Y”, etc.
Yes, but Recall is spyware by design posing as a benign feature. This kind of unethical behaviour I vehemently oppose.
The best way is to use comms channels that avoid their Windows install entirely. If Recall never sees it, it never gets recorded.
If there’s anything sensitive I’m communicating with someone digitally, I make sure that the person in question has basic tech security skills and knowledge about privacy, including telling them to stop using Windows. Including taking the time to teach them basic stuff (like full disk encryption, VPN and Tor usage, explaining E2EE, etc) myself. If you have a high threat model but are talking to non-techy people, you should be taking the time out of your day to do this.
If you’re thinking “wow I can’t be bothered to do all that”, your messaging is probably not sensitive enough for this to be a significant concern. Not that “if you have nothing to hide you have nothing to fear”, but just “the amount of time you put into security and privacy should be proportionate to your threat model and the cost of compromise”.
Either use secure, encrypted VoIP calls (e.g. over Signal or another secure messenger with an end-to-end encrypted call feature)
Or you use a secure messenger that only runs on smartphones and doesn’t have a desktop client
deleted by creator
This is just horrible, fuck big tech and their services
Absolutely
You must start spreading libre software effectively. You don’t control their device. You must show them how to fix it.
Turn off your computer, move to a cave in the mountains, and abandon society.
A bit extreme but there is nothing you can do to stop your messages from appearing on Windows machines except not sending them to anyone who might view them on Windows machines…which will definitely be nearly impossible in 2024
I couldn’t wait to post this obligatory fragment of Parks and Recreation - Ron vs. Online Privacy: https://youtu.be/8xn1rO1oQmk
Works great until some hikers take a photo with you in the background, that gets backed up to iCloud, then they want to show the photo to a friend, download it to their computer, open it and BOOM, Microsoft AI knows your face
I’m in Europe :p
who cares? try to prove anything
Europe has decent privacy laws, that’s how I avoid it
I’m afraid this comment shows a severe underestimation of the gravity of the issue. Windows recall doesn’t stop at borders even if it were illegal there.
Well, it’s not here yet. And I do use windows 11, as does my mom, my grandparents and other pc’s I’m the one helping with. I don’t recall any recalls :p
And if they do push it here, it’s probably followed by a news headline “eu fined Microsoft 10 billion for gdpr violation” or something like that
EU fines take way longer then that, give it a couple years of data collection and if we’re lucky they get fined
Their warnings tend to be a bit quicker
It can be turned off so it’s up to the person you’re messaging. Once you send something the person at the other end is in control of what happens to it.
Once you send something the person at the other end is in control of what happens to it.
True, but this is the beauty of trust. I decide to communicate one way or another with someone depending on the level of trust. Them deciding to break that trust is a risk I chose to take. However, I do not choose to communicate with Microsoft, whatsoever. Windows Recall is the most blatant piece of spyware ever; beyond comprehension how this is so normalized.
Society just needs to get over this AI fad atm. By which I’m not trying to say that AI won’t revolutionize pretty much everything in our lives eventually, but first we need to figure out what it can actually be useful for. Or rather non-tech people need to be fully introduced to both its benefits and its pitfalls before tech companies will have a clear picture of where the red lines are for people ideologically speaking. We the nerds have our moral compass figured out but we’re a minority when it comes to who these products are made for.
Leave it to Microsoft to come up with the most dystopian AI concept yet. But to be honest I’d be way more wary of a company like Alphabet for whom data collection is much more central to their business model and who know how to package their spyware neatly. Microsoft announcing this as a feature from a podium shows how tonedeaf they are but I’d argue it also shows that they’re not following some self-serving plan behind the scenes to take advantage of that thing they’re so proud of publically (a mass espionage at which I firmly believe they wouldn’t be anywhere near efficient enough if they tried). They really must’ve thought that this is what can get Windows back into the limelight. It is Microsoft’s problem of our time that with everyone being on smartphones and tablets now they are losing traction in the consumer market by the day.
Point being (as far as the valid privacy concerns go) that Microsoft were never in the data business. They’re just really really bad at understanding what consumers want out of an operating system. I got my first own PC in 2001 right when XP came out. They’ve always been bad at making things work for the user. And since Vista all they’ve really been doing is copying Apple’s eyecandy. First off of macOS (then OS X), now with Windows 11 they basically want to look like a tablet OS with app icons once again after that idea failed spectacularly under Windows 8. I’m basically just rambling at this point but it should go to illustrate their lacklustre corporate decisionmaking. I wouldn’t be worried about their potential desire much less their ability to compromise that Recall data. Yes it’s a hugely concerning concept from a privacy standpoint and every step to circumvent its analysis should and arguably must be taken, but I also wouldn’t lose sleep over the data it is collecting on other people’s machines.
You have to trust the person you’re communicating with has turned it off. That’s my point. It’s an optional feature
Then you have to trust the person you are communicating with has turned off windows recall. That has to be the starting position.
Tools will come to block or break windows recall but it will still be based on trust that the recipient is using them. Privacy centred apps like Signal wouldn’t want windows screen shotitng every message for example. There are many apps and tools including in the professional sphere that would not want their data leaking via recall so it will come.
Unfortunately it may come late in the professional realm probably after scandals break. Employers using recall data to investigate staff for example - it’s bound to happen eventually.
My own organisation, a huge health organisation, has opted in to CoPilot. It’s crazy in my view, even if our data is ring fenced in some way. I don’t want private patient information being used to train Microsoft shitty tools, or stored on their servers. Regulation and the law is way behind when it comes to this stuff.
Wow, valid issue.
Spitballing, potentially a secure app could run memory only, blah, blah, blah. Nope, you’ve given M$ your screen FFS, it’s all over. If you care, move elsewhere, tell your friends…
As you point out, codes are an option, but it’s not a slippery slope, it’s a waterslide.
If the content CNA be displayed, it can be parsed by recall.
The only way I can see to bypass it is to obtain DRM keys and display your content on a website only if widevine is active, like Netflix does. Surely it can’t screenshot DRM protected content, but also this is Microsoft .
Lol what if microsoft accidenttly gives people a way around that. Want to screenshot something from netflix? Just watch it then look where recall stores all the screenshots!
If you tell something to someone else, assume it’s compromised.
“Three can keep a secret if two of them are dead.”
(Even then I’m not so sure)
Don’t forget that while they managed the PR better, apple “Intelligence” also has access to damn near everything on your devices.
Yet there’s no backlash because they’re not so stupid as to say “we’re gonna take screenshots as you go so we can improve your digital life kthxbye”.
“Private cloud” as if that isn’t an oxymoron.
People also willing submit to the cult if Apple and just believe everything they say. People are likely more frequently forced to use windows due to work or just the lack of choice for less technically confident people.
If it leaves your device, you cannot control it.
Right but you could at least be reasonably sure it wouldn’t be outright spied on from the person you’re sending it to. Now it’s almost a guarantee.
Like if I sent something to a friend of mine, I could be fairly certain it wouldn’t end up in the wrong hands unless they got compromised or did something stupid. I could trust their competence.
Now everyone that isn’t actively managing their own windows installation is absolutely compromised, as a rule. Like I can’t just send an email to my mom anymore, from now on its always my Mom and Copilot.
So is there a way for businesses to disable this garbage feature through managed device settings or something? I’m guessing corporate legal departments aren’t going to be too thrilled with this feature.
There’s a CSP for disabling it on windows enterprise devices at least. Not sure if there’s a way for pro and home machines.
There must be. Recall and info sec is mutually excluding by definition!