@ranx
Thanks to the use of good-guys-only backdoor, because that's totally a possible thing that exists. Yeah, right. 🙄

#chatControl

But you expect us to believe there'll be no scope-creep?

What about the "AI" scanning of all messages on the client-side, whose security impications are so bad that all politicians and all military personnel are exempt?

This is mass surveillance. Firstly, it's an enormous treasure-trove of data, that will inevitably be breached.
Secondly, abuse of it by those controlling it is inevitable. And even in the unlikely event that it isn't actively used for identifying and suppressing dissent, the prospect will cause people to self-censor.

The Stasi would have wet themselves at the possibilities.

Every subject-matter expert on every aspect of this has called it out as horrifying, and pointed out comprehensively that it cannot and will not accomplish its stated goals.

Its supporters are AI and surveillance enterprises who stand to gain financially.

So let me be clear: there is nothing necessary or proportionate about this. The only fair resolution is to scrap the whole proposal.

@ranx I'm happy to see you respond to critical comments!

You can't detect illegal materials inside a message without accessing the message contents. Whether you AI-scan the message contents before they are sent or break the encryption afterwards, it will have very negative consequences.

There are many govt and non-govt organisations protecting children and they tend to be woefully underfunded. Please consider funding them instead of focusing on broken ideas.

@ranx the only way to detect the sought content is to collect and analyse all content.

The carve out for politicians makes it obvious.

as a little privacy advocate let us break it down one by one

"Let us be clear: under this proposal, there is no general monitoring of online communications. There will be no such thing as “chat control”."

wanna know what the definition of insanity is? Step aside far cry 3 character, we got a new champ on the block (1/4)

That might work on evenint news, but not around here.

Phony shit. Just what we‘re used from your fine president.

@ranx

"Only material that is clearly child sexual abuse will be searched for and can be detected."

Whoever is responsible for this “response” deserves a raise for their outstanding achievements in bullshitting Europeans. The above quote sounds benign and harmless, but in reality it means the following:

Every message *may* contain child abuse content, therefore every message will be scanned.

@ranx

No, YOU be clear. You are going to use AI for so-called "child protection". Which hell on earth in-progress did you take inspiration from? USA? Russia? UK?

You have no spine. Grow one.

You clearly have no understanding of IT, encryption, detection algorithms. Your proposal will lead to massive false positives, privacy violations, and is a great basis for mass surveillance.

I want a strong Europe which protects freedom of people, improves prosperity of all Europeans, protects our environment, and protects our democracy.

However if Europe goes the way of this proposal, I'll become a strong and vocal opponent of the EU. And I won't be the only one.

@ranx

@ranx IT is not magic!

If you want to detect whether a given message contains CSAM you obviously *have* to let it be scanned by some detection algorithm / machine learning. Otherwise you simply can't know...

In other words *each and every* message would have to be exposed to the threat of false positives. It is completely outrageous that you try to pretend otherwise!

#chatcontrol

@ranx

I have a question about: "Only material that is clearly child sexual abuse will be searched for and can be detected."

How? How will one determine when it's "sexual abuse" on all media? One has to scan each and every single media file to find these, I do not see any other way

Don't get me wrong, I would love a better way to deal with CSAM content but scanning all to find the few is not the way

@ranx

> under this proposal, there is no general monitoring of online communications

There will be a monitoring of every single picture and every single link in every single piece of online communication. Every message will need to be scanned for the presence of links and images. That's very different to "no general monitoring".

What you describe is technologically impossible.

Source: Me, cybersecurity professional.

@ranx

Hey! A massive thank you for engaging here. That is wonderful, and everybody involved is beautiful 😀

On the chatcontrol topic, unfortunately the messaging is not correct. The whole IT industry is very concerned (as we know, that is the highest level of concern) about chatcontrol for a good reason. We know it is impossible to implement it in a secure way that is not abused.

How could this message best be relayed to the commission that does not involve harassing them in person?

What you should be doing instead is to make it both easy and legal to report CSAM. For example, people moderating Mastodon instances hosted out of Germany must immediately delete any CSAM the find without reporting, because if you report, it's proving "possession" which comes which a 11 year minimum prison sentence.

What you need to build up is easy reporting like the USA has, and fund sufficient staff to actually deal with the reports.

That's not what "searching" means.

Here, let me help:

merriam-webster.com/dictionary…

Definition of

Definition of '' by Merriam-Webster

^{Merriam-Webster Dictionary}

@ranx
It doesn't matter how you'll call or phrase it, it's still mass surveillance. There are better alternatives to protecting the children such as having more undercover personelle who are specialized in infiltrating and disbanding child trafficking rings thus saving more children and many better options than to invade anyones privacy for it.

Also, we all know for sure that AI's have a tendency to be severly trigger happy. In other words, it's doing more harm than good.

Let us be clear: you are aiming exactly for #ChatControl.

• You know exactly that encryption cannot be "selectively" broken.
• You know exactly that you want to establish technological infrastructure that can search for anything - CSAM is merely one of a myriad of use cases you're already drooling about.
• You know exactly that your blahblah on due process does not convince even a fly on a window pane.

Just keep your lies and propaganda to yourself – nobody believes you.

@ranx

look EU commission, there is no magic way to only scan for CSAM, all messages will be scanned.

Here's a document about cryptography and backdoors written round about the time of the clipper chip. TLDR:

A backdoor is a backdoor.

datatracker.ietf.org/doc/html/…

#ChatControl #NoChatControl

@ranx

oh sure. because secure, accessible only to good guy backdoors work great and we're totally on point and accurate in not looking at any files that aren't "bad".

let's get the starship enterprise food replicator to whip up a nice chamagne we can toast this with, since that technology is fully as possible as the above.

The current proposal is in practice just the old "chat control" proposal wrapped into a new wrapping. It carries the same flaws as all the other proposals.

The common denominator is that there is a belief in technology to provide the tools to enforce a policy for the good of the people.

But technology is just that - a tool. It is can and will be abused. There is no technology available which will only protect us. In the wrong hands, it will be used to do bad things.

Just as a hammer can be used to build wonderful buildings - but it can also hurt people. And it will hurt people, even if there are laws forbidding such use.

Replace "hammer", with "weapons", "cars" ... even "computers". It will always be possible to abuse it to do unintended things.

Technology cannot solve issues related to how we as citizens relate to laws.

Further, technology cannot differentiate between "good" or "bad". Just as a hammer cannot differentiate how it is being used. Only the moral of the person using the tool can differentiate if the tool is being used as intended.

Please try to understand the analysis from Patric Breyer.

patrick-breyer.de/en/posts/cha…

A short video explaining the issues as it currently stands: digitalcourage.video/w/xziZATY…

Chat Control: The EU's CSAM scanner proposal

🇫🇷 French: Traduction du dossier Chat Control 2.0 🇸🇪 Swedish: Chat Control 2.0🇳🇱 Dutch: Chatcontrole The End of the Privacy of Digital Correspondence Take action to stop Chat Control now! The Chat Control 2.

^{Patrick Breyer}

@ranx

Wow. Watching as an outside observer, I'm still struck how none of the @EUCommission words above are true, nor defensible. Why the hollow talking points absent good faith engagement with the real issues?

Sorry to be blunt, but that answer is ridiculous.

In order to be able to identify illegal material, you WILL have to search EVERYTHING on the phone (or computer) of every citizen in the EU.

How else would you know which picture, movie or document is illegal and which isn't?

That is mass surveillance the way Russia, China and other totalitarian states behave.

As you do this without any probably cause, this violates the most basic principle of a proper jurisdiction.

@ranx
"Only material that is clearly sexual abuse will be searched"

That is such a blatant and clumsy lie! What do you take us for? What have you been smoking?

How do you suppose to find such materials if not by searching through everything?

If you already know that some material is abuse, what need is there to search? You already have it.

When you're trying to find a needle in a haystack, you will have to dig through all the hay.

@ranx

Exactly like luggage scanning at the airport. Only luggage containing forbidden substance are searched. No mistake, all perfect.

Of course not! At the airport, everything is scanned, mistakes happen, agents confusing normal things for dangerous things. People are being pulled aside, interogated, sometimes for no reason (random control).
Carrying too much liquid is criminalized.

People are blamed for criticizing the system that "makes everyone safer".

@ranx

@ranx “Only material that is clearly child sexual abuse will be searched for and can be detected.” IF YOU WANT TO FIND ALL, YOU HAVE TO SEARCH ALL

I do not usually shout, but now I feel I must. Sorry

@ranx

Thats a lie and either you are fully aware of it, or you actually believe in unicorn technology that don’t exist. Both options scare me equally…

That's just not how computers work. Making restrictions procedural rather than technological is a dealbreaker.

This response actually makes me more anxious in that it shows either the issue is not understood or is being deliberately misrepresented.

All you were doing with this "Chat Control" proposal is trying to make Europe a surveillance state like America already is. Please learn from America's mistakes. You also set an extremely dangerous precedent by passing something like this.

As an American I am begging you to PLEASE not set this precedent and make yet another place a surveillance state.

@ranx

How exactly will you "search" and "detect" without scanning everything? Technically this is not possible and everyone knows that.

The question is: How do you know the data thats "clearly child sexual abuse" from the other? You can not decide that without scanning the other contents too.
And how about innocent pictures that I take of my child in summer? Will you look at those to find out if they are pornographic? I don't want anyone to look at pictures of my child without my authorisation.
Especially not totally strangers who might have a fetish themselves. How should I know? Or even worse: Some AI that can not differentiate between innocent beach pictures and real documentation of child abuse? You will have to take a proper human look on so many private family pictures. So actually you are putting parents an families under surveillance-pressure. How dare you?
And why should I trust you, if you have not even understood, what you are doing there?
And also: How can we protect EU law from antidemocratic forces, if you use it against us?

Thanks for being in the fediverse and even answering to questions here, though.

Hello @EUCommission!
Let me be clear: under this proposal, there is absolute general monitoring of online communications. There will totally be such a thing as “chat control”.

Not only material that is clearly child sexual abuse will be searched for and can be detected, but every single media!

@ranx As so many have already mentioned, this is a blatant lie. You need to scan *everything* to kick off that "thorough process", violating privacy and freedom of your citizens - us - to begin with!

Even if this isn't abused in any way, any false positive or dubious accusation will ruin someone's life. At this scale, Chat Control has the potential to ruin thousands of lives, if not millions worldwide. Even 99.9% accuracy simply will not do!

@ranx You know what's downright scary? That people who have so little understanding of how the technology works making claims like that.

There will be an extreme amount of false positives, meaning random people will be looking at my private photos for no reason at all.

This will absolutely not help with CSAM, but it will give police a legal framework to watch people.

@ranx

This is also highly related to this discussion:

axbom.me/objects/046a5394-7d92…

Per Axbom (@axbom@axbom.me)

The people behind #ChatControl claim that they are not breaking end-to-end encryption. At the same time, private messages on services with end-to-end encryption will be intercepted to uncover crime...

^axbom.me

Sensitiver Inhalt Zum Öffnen/Schließen klicken

@ranx I wonder whether seriously not understanding or lying about it would be worse here.

What ChatControl does is like opening and reading every single letter being sent. Yes, maybe you only take photo copies of naked kids, but you do read all the letters. That's the opposite of freedom and every dictator's wet dream.

as someone who does content moderation for a living, I like to think that I know what I’m talking about.

There is so much more nuance than people realise when it comes to CSAM. It’s already a challenge for human reviewers, especially at scale. As an (over simplified) example, how do you plan for an LLM to tell the difference between an innocent photo of a naked toddler sent by a parent vs a similar photo but only slightly modified with the intent to sexualise the minor?

You also have the issue of accurately identifying the age of a potential victim. A baby is easy enough to identify as a minor, but as you get closer to 18 that becomes a lot more challenging.

This is all over simplified and skips over other issues, but, the point stands. You CANNOT detect “only material that is clearly child sexual abuse”. You WILL end up looking at false positives, and it WILL be far more common than you think/plan.

@ranx

So you don't monitor chat, but somehow you know when it contains something illegal, sure....