News for hackers. Stuff that matters.
Re:Kyoto Statement on End-to-End Encryption by SenAnder on Friday October 20, 2023 @12:47PM (#37958489)
The key to winning the encryption fight is not in nuanced arguments, but to call anti-encryption/client-side scanning laws what they are: mandatory submission of all private communication to government inspection.

Because ultimately that is what those laws serve. Targeted surveillance can already be accomplished by planting a bug on the suspect's hardware, or pointing a telescopic lens/hidden camera at their monitor/keyboard. You only need encryption back-doors if you want to surveil in bulk.

Re:Kyoto Statement on End-to-End Encryption by charcircuit on Friday October 20, 2023 @12:51PM (#37958531)
You don't need a backdoor. With a responsibly designed cryptographic protocol it would be possible. You could make cryptographic protocol where all chats have an extra user in them whose key is unable to be decrypted or revealed unless there is a warrant provided.
Re:Kyoto Statement on End-to-End Encryption by AlexandrB on Friday October 20, 2023 @01:00PM (#37958635)
How is what you're describing different than a backdoor in a practical sense? With this setup the privacy/security of the system now hinges on how carefully the company (and possibly law enforcement) secures these "3rd party" keys. A bad enough compromise/leak could make everyone's chats accessible overnight.

Edit:

> unless there is a warrant provided.

This element is also difficult in practice. What stops a rogue employee from taking a peek? Now you need to have strong organization controls on access, audit trails, etc. It expands the problem of private, secure chat out of the realm of "is this encryption system secure" and into the realm of "is this organization managing and auditing data access responsibly". The latter is a much harder question to answer and more prone to change over time with no notice.

Re:Kyoto Statement on End-to-End Encryption by SenAnder on Friday October 20, 2023 @01:24PM (#37958936)
> unless there is a warrant provided.

> This element is also difficult in practice. What stops a rogue employee from taking a peek?

More pertinently, a warrant is not a magical guarantor that the surveillance is legitimate (nor can a backdoor be made that only works with a warrant - a warrant is not a technical barrier). If a government is already so anti-privacy that they force all their citizens to use backdoored communication, we can expect warrants to be rubber-stamped. Eventually some secret court will approve a warrant that covers an entire city or country, and it'll be renewed every two years as a routine matter.

Re:Kyoto Statement on End-to-End Encryption by charcircuit on Friday October 20, 2023 @04:50PM (#37961346)
Keys can be stored in hardware meaning that they need to be physically stolen. The keys don't need to be owned by emlpoyees. You could have them assigned to judges. The hardware keys can have rate limits so stealing a judge's key has limited impact before the key is revoked.
Re:Kyoto Statement on End-to-End Encryption by tzs on Friday October 20, 2023 @10:16PM (#37963499)
Actually technically those issues could be addressed using encryption.

Use a threshold scheme to share the warrant key among multiple independent share holders. Use enough share holders and a high enough threshold to make the odds of enough share holders going rogue or being corrupted to reveal the warrant key lower than the odds that someone will brute force the key.

In the limit you could make everyone in the country a share holder and require a large supermajority for key recovery.

Re:Kyoto Statement on End-to-End Encryption by SenAnder on Friday October 20, 2023 @01:26PM (#37958963)
> With a responsibly designed cryptographic protocol

Facilitating mass surveillance of populations is the opposite of 'responsible'. It's what allows tyrannical governments to remain in power.

Re:Kyoto Statement on End-to-End Encryption by charcircuit on Friday October 20, 2023 @04:54PM (#37961390)
Warrants are not supposed to be used for mass surveillance. Rate limits can be implemented into the design to avoid that from happening.
Re:Kyoto Statement on End-to-End Encryption by SenAnder on Saturday October 21, 2023 @03:15AM (#37964773)
Yet the US had (has?) a secret court that signed off on PRISM. Without strong technical barriers in place, such as ee2e, such things are near-inevitable - surveillance is just too easy. Like putting a pile of cash on your lawn with a sign saying "Do not steal". You can dream of some ideal society where laws are so well respected and enforced that nobody steals the cash, but experience suggests, for the pile of cash and for mass surveillance, that you will be disappointed.
Re:Kyoto Statement on End-to-End Encryption by gray_charger on Friday October 20, 2023 @01:17PM (#37958847)
> You could make cryptographic protocol where all chats have an extra user in them whose key is unable to be decrypted or revealed unless there is a warrant provided.

...that's called a backdoor.

Re:Kyoto Statement on End-to-End Encryption by charcircuit on Friday October 20, 2023 @04:53PM (#37961365)
A publically documented part of a system that explicitly states its purpose is not a backdoor.
Re:Kyoto Statement on End-to-End Encryption by Tyrannosaur on Tuesday October 24, 2023 @02:03PM (#38002880)
A publicly documented part of a system that explicitly states that its purpose is to compromise encryption and privacy may not technically be a "backdoor", but that's a distinction without a difference.

https://www.justsecurity.org/16503/security-front-doors-vs-b...

Re:Kyoto Statement on End-to-End Encryption by akerl_ on Friday October 20, 2023 @06:33PM (#37962225)
How does the extra user's key know not to reveal itself unless a warrant is provided?