Hacker News Reader

Comment by gickhocq | Hacker News Reader

I don't really see the problem with this. Scanning happens before the message is encrypted or after the message is decrypted (thus preserving the end-to-end encryption itself from compromise), occurs only on the device itself, and will report back only if known illegal content is detected.

It seems to be the best possible method of implementing such scanning. Shouldn't we be supporting this as a proportionate safety measure?

View on HN

What's been implemented is a warrantless search on every file. The only real restriction is that you have to know what you're looking for. Right now it's looking for this one thing 'for the children', but it could be used for anything. What if a law is passed requiring it to look for copyright marterials, or unauthorized chemistry books? What if Apple starts searching for say, leaked source code, or some an app they have a patent dispute with? What if Apple does nothing else and keeps it restricted to what they say it will be restricted to but it becomes more widely adopted by even less ethical manufacturers?

Any hole is a wide hole.

And to think that wide hole isn't now utilized is folly, despite whatever is said.

Known illegal content will inevitably become possibly illegal content under that scheme given the progress of AI. It will lead to all kinds of negative (and even fatal) consequences given the tendency for parents to have photos of their toddlers bathing and involvement of LEO. Parenting is hard enough without the risk of being murdered by a cop.

Once it is possible to scan the messages, you can bet it won't be limited to CSAM, that's just to get their foot in the door.

As of today every tech platform scans every message, photo, email, etc. for CSAM, yet there isn't misuse of the system. Why would that change under the proposed change?

Is it just csam? Really? They don’t search things for anything else? Not at all? And what if iran tells them they must scan for something else? Obviously they’ll refuse

1) Iran? did you forget about sanctions? Apple does not operate at all in Iran. No offices, no employees, no data centers. Nothing. Apple doesn't care what Iran tells them to do lol.

2) I mean they search for spam, malware, copyrighted material (as required by the DMCA), etc.

Define tech platform.

It's proportionate to give partisan governments fond of the nastiest behaviors the ability to read every word that you type to anyone?

Even now, iOS and MacOS have scanning ability that should be assumed is used well beyond the proposed scope.

Anyone who uses these OS's can only come away with the sense that literally everything is logged.

> It's proportionate to give partisan governments fond of the nastiest behaviors the ability to read every word that you type to anyone?

This argument is so absurdly dumb. Today, every tech company will give EVERY single message, text, IP address, photo, email etc. to law enforcement. Yet you rarely see misuse of this system. Why would that change under the proposed rules?

If your argument is "trust them bro", then I'm not the one who is making "absurdly dumb" statements.

N-gate summed it up very well: "if a guy in Cupertino can put a U2 album on your phone, he can just as easily put CSAM on it".

If you are using a device managed by Apple/Google/Microsoft, E2E is merely a roadbump - it means that your device owner needs to push a backdoored update to it, rather than simply reading your data server-side at their leisure.

Which is still valuable, to be clear: it's a pretty large roadbump. Either you're such an important target that they're shipping a specific update to you alone (in which case, frankly, you didn't stand a chance), or they risk exposure.

But if you install a little snitch that reads your messages and reports you if they match some complicated, necessarily hashed pattern, then that roadbump is mostly gone. All the guy in Cupertino now needs to do is update the patterns to match what they want them to match, without anyone noticing.

I think I see your argument, are you saying that because the scanning involves transforming the content into a hash and comparing to a set of hashes, the problem is that this is resistant to third-party analysis? Not because of the overall behaviour of the scanner or indeed any other software on the device, which can be reverse-engineered using standard techniques, but because a database of hashes is inherently opaque as to the content it will match with?

Correct. Is that hashed fingerprint you just downloaded associated with the face of a known victim of abuse, or with that of a whistleblower? Impossible to tell.

(The N-Gate scenario is a little different: someone who doesn't like you controls a random app that is allowed to write to your /Downloads or /Pictures folder. Puts compromising material in it, then lets the well-meaning OS scanner pick it up and report you. You spend the next X years wrangling with law enforcement, trying to convince them that you were "hacked", the same thing every criminal claims.)

If the scanner program merely refused to transmit the potential match, it would be acceptable. No avenue for abuse (without shipping a malicious update, which is always possible for cloud-controlled devices anyway), and at worst an annoyance in the event of a false positive.

But I very much doubt that the parties pushing for this legislation would settle for that.

Reply on news.ycombinator.com