Another case of setec astronomy.
Tainting the CSAM client-side scanning database
21–30 of 276 posts
Re: Tainting the CSAM client-side scanning database
#22The only thing that can save us is a huge number of false positives and a backslash on the governments in charge, when they'll have to concede for the 10,000th time that it was not CSAM but parents sending their baby pictures to the grandparents.
Re: Tainting the CSAM client-side scanning database
#23The issue described here, to my understanding, is that you find or create csam and then manipulate it so that its fingerprint collides with another image that you want to be flagged as csam. You then submit the manipulated version of the found or generated image to the authority. First, at what point does the authority go "uh... Where did you get this from?" Practically speaking, the people doing this would have to b…
Would this org refuse to accept CSAM from an anonymous source just because they can’t point to its origin?
Re: Tainting the CSAM client-side scanning database
#24How do these databases differentiate between AI generated CSAM and CSAM of real victims? (Since many jurisdictions only criminalize real CP)
I know that 99% of people cannot tell an AI image from a real photo since that "Last giant irish greyhound 1902" photo has been going around on social media for weeks, and it is, to me, unbelievably obvious AI.
Re: Tainting the CSAM client-side scanning database
#25Any computational method that relies on a function that converts m bits (an image) to n bits (a fingerprint) where m > n will always be vulnerable to such an attack. The smaller n is compared to m, the easier it is to counterfeit something with that signature. It is not new knowledge. The only way to be certain, unfortunately, is for a human to look at it. When the allegation is as serious as CSAM, I would rather be…
> Any computational method that relies on a function that converts m bits (an image) to n bits (a fingerprint) where m > n will always be vulnerable to such an attack.
The "vulnerable" depends on your definition: SHA-2 and SHA-3 are both still quite safe against preimage attacks, and even second preimage attacks require significant work to pull off for SHA-2, and I am unaware of any meaningful second preimage attack on SHA-3.
Of course, SHA isn't built for finding similar images, but for finding exact matches this should be safe enough.
Re: Tainting the CSAM client-side scanning database
#26[flagged]
Can you point to any society anywhere, where people learned to behave?
Re: Tainting the CSAM client-side scanning database
#27Because client-side scanning is not going to work, and no one wants to government issued black box binary to send their conversations and photos to unnamed police person randomly, the non-compliance is the only way. People just start to use chat programs in the EU that do not comply. This would be Signal, Telegram, others. The EU can fine and fight with Meta/WhatsApp, Apple, others, but that’s about it. The EU bureau…
Signal advises its users to install the app from the Play Store. While you can still get the .apk from Signal’s website, the developers warn against that. The EU can definitely exert pressure against what is hosted on the Play Store. Telegram famously lacks end-to-end encryption (unless you intentionally use its private-chat feature, which few people do) and shouldn’t be mentioned in the same context as Signal.
If you enable private chat, you don't get notifications for new messages, that way the feature is self defeating, it seems to me at least. (Something could also have gone wrong?)
Re: Tainting the CSAM client-side scanning database
#28Because client-side scanning is not going to work, and no one wants to government issued black box binary to send their conversations and photos to unnamed police person randomly, the non-compliance is the only way. People just start to use chat programs in the EU that do not comply. This would be Signal, Telegram, others. The EU can fine and fight with Meta/WhatsApp, Apple, others, but that’s about it. The EU bureau…
Signal advises its users to install the app from the Play Store. While you can still get the .apk from Signal’s website, the developers warn against that. The EU can definitely exert pressure against what is hosted on the Play Store. Telegram famously lacks end-to-end encryption (unless you intentionally use its private-chat feature, which few people do) and shouldn’t be mentioned in the same context as Signal.
Molly is a just as good (if not better) fork of Signal that is distributed on FDroid. It's the same network and same chats/contacts, just a different front-end.
And Signal avoids FDroid because they don't want someone else signing packages but they can always provide an FDroid repository like many others do and sign everything themselves.
If push comes to shove they'll be fine and pressure to black box signal in the EU is unlikely to hold up if they can just move users to another app store.
Re: Tainting the CSAM client-side scanning database
#29Earlier quoted context omitted.
Signal advises its users to install the app from the Play Store. While you can still get the .apk from Signal’s website, the developers warn against that. The EU can definitely exert pressure against what is hosted on the Play Store. Telegram famously lacks end-to-end encryption (unless you intentionally use its private-chat feature, which few people do) and shouldn’t be mentioned in the same context as Signal.
If you enable private chat, you don't get notifications for new messages, that way the feature is self defeating, it seems to me at least. (Something could also have gone wrong?)
Notifications generally go through Google Firebase so by not running private messages through firebase they avoid potential leaks. At least that's my guess
Re: Tainting the CSAM client-side scanning database
#30The issue described here, to my understanding, is that you find or create csam and then manipulate it so that its fingerprint collides with another image that you want to be flagged as csam. You then submit the manipulated version of the found or generated image to the authority. First, at what point does the authority go "uh... Where did you get this from?" Practically speaking, the people doing this would have to b…
Another version of the attack is to manipulate an innocent image in a way that's flagged by detection systems, and then spread that image somehow. If any of the responses are automatic, then it's a lot of false positives in the system, a lot of contact from police, disabling of the host, things like this.