[flagged]
Are you kidding me? How do you think predators get in contact with children? Over social media that children use, obviously.
And of course many criminals use chat. Most have no clue about encryption, like the vast majority of humans.
I thought it was pretty common knowledge that the vast, vast majority of the perpetrators of these offenses are either family members, or known and trusted people to the family, such as the friends of an older sibling, friends of parents, stepparents, teachers, priests, etc. The bogeyman of the evil stranger paedo lurking on social media cold-calling random kids is an edge case.
I thought it was in fashion. Happens where I live.
Well, the point is to scan all messages, period.
And then to detect those that come from predators, not adults. How often do parents convince their children to send... revealing pictures? Or to meet them somewhere? How often do parents introduce themselves to their children in messages?
You can't seriously believe that a conversation between parents and children always looks like a conversation between a predator and children, can you?
YES. That's the problem. Whoever controls it has that power. We don't want that. That's the argument against ChatControl: "imagine that those who get in power are against you, and imagine what they can do if they abuse this tool".
But saying that "a law enforcement officer may see a false-positive between a parent and their child and I find this insufferable" won't convince many people, IMHO.
Seems like a terrible idea.
I am saying this:
> You can't deny that if you can read all communications, then it's easier to detect CSAM than if you can't even see it.
I am against ChatControl, but people who say "it shouldn't exist because it is useless because it cannot ever help detect anything illegal" are wrong.
It doesn't help the cause to flood people with invalid arguments.
You can see this clearly with the constant inflation of AML laws in the EU, which become more and more restrictive and invasive each year, without any clear effect.
It has started before Trump, I think that a turning point was the Patriot Act, but Democrats didn't overturn it and picked their ennemies, too[0].
But you can't. You can legally require messages to implement client side scanning before the encryption happens or add some backdoor keys and hope they don't leak.
Since cryptography is known anybody can always just create messengers that just don't implement that requirement. If you sexually abuse children the hurdle to illegally running a non-backdoored messenger seems pretty low to me.
Thats like fighting illegal street races by implementing electronically-GPS-enforced tempolimits for everybody. You won't catch the people it is meant to catch. Only that surveilling your entire population minus the criminals has dire consequences for a free society.
Be careful what you wish for. We could enforce client-side scanning on the OS. Everything that appears on the screen is scanned and reported.
> If you sexually abuse children the hurdle to illegally running a non-backdoored messenger seems pretty low to me.
How do you contact children on mainstream messengers if you can't use mainstream messengers?
Not to mention that most people consuming CSAM are not technically savvy. It is known that such material goes through mainstream messengers.
> Only that surveilling your entire population minus the criminals has dire consequences for a free society.
Again: I am against ChatControl. We fundamentally cannot know what is running in this client side scanner, and that makes it a risk.
But saying that it won't ever catch a single criminal is not a valid argument. You won't convince politicians to vote against ChatControl with such claims.
The reason is because it works. They're not stupid - they can use signal.
The reality is that the privacy options not only exist, they're really good - often better and easier to use than the mainstream stuff.
They will just pivot to other tools.
My understanding is that they are not only talking about having a list of hashes (of illegal images), but also having some kind of machine learning. And they are not only talking about scanning images, but text, too.
I don't know what you expect them to report when scanning conversations with machine learning?
Nope. Criminals can still just build their own devices with their own operating systems. We have existing OS without spying, people have them as ISO on their harddrives. You can't stop criminals from installing an old Lineage OS build.
Legally you can't stop anybody from using encrypted channels if they are motivate unless you go out and erase the knowledge.
Implementing filters that go on all communications is undemocratic. Any future authoritarian government can use the same filter to immediately target their opposition, not just in theory, in practise. We have designed our democracies with division of powers for the simple reason that we have learned through history that giving any single kind of entity that kind of power leads to tyranny. That means whenever we give the government new superpowers we are changing a carefully setup balance. What does the citizen get for that trade? Nothing. If your power only works as long as it is in the hands of the good ones and becomes dystopian once it gets into the hands of the bad guys, maybe that power shouldn't exist.
Since we want to obviously prevent childrem from being sexually abused the best way to start according to most child-protection organizations is to start at the root. That means educating kids early on in child-suitable ways and investing in prevention and psuchological help for potential pedophiles. If children have the possibility to signal such a thing happened to adults or other persons of trust, you don't need mass surveillance.
But my guess is that CSAM is just the symbolic reason, in reality this is meant to do more. It would be perfect to target small movements as they are emerging for example.
Haha sure. You over-estimate many of them. Not everyone breaking the law is a professional criminal. Especially when it comes to CSAM.
> Legally you can't stop anybody from using encrypted channels if they are motivate unless you go out and erase the knowledge.
ChatControl doesn't pretend to do this: they pretend to control mainstream social media, whether encrypted or not.
> It would be perfect to target small movements as they are emerging for example.
Yes, this is my problem with ChatControl: it's a powerful tool of surveillance that would be very dangerous if abused. We don't want to create that.
But "it's useless because criminals will compile their own kernel anyway" is an invalid argument. It doesn't help the cause.