3 ms·
This is very emphatically not an answer. Iran learned this the hard way - their centrifuges were isolated from the internet, but someone managed to get access t
by makomk 11y ago
This is very emphatically not an answer. Iran learned this the hard way - their centrifuges were isolated from the internet, but someone managed to get access to the memory sticks used to transfer data in and out and inject an exploit that compromised the critical machines the moment it was inserted. At some point you inevitably need to get data in and out, and that means you need a secure device. (For that matter, how are you loading software onto those devices in the first place? At some point, the software development and provisioning process will have contact with the internet.)
- smegel 11y agoThe contrarian view is there is no such thing as a "secure device", and physical security is the only answer.
- chimeracoder 11y ago> Iran learned this the hard way - their centrifuges were isolated from the internet, but someone managed to get access to the memory sticks used to transfer data in and out It's not a sufficient answer in itself, but it's a good response. It's a lot harder to compromise a physical system than to compromise a network. Furthermore, it's a lot harder to compromise a physical system and go undetected than to compromise a network. Both are possible, but given an offline system with good physical security and an online system with good physical security, you're still better off with the offline version.
- zwarag 11y agoYeah right like. What system cannot be compromised when you have physical access to it? You think if they ran openBSD or Windows or Apple Server, that they would not been hacked? All systems are exploitable with [BadUSB](https://github.com/adamcaudill/Psychson https://github.com/adamcaudill/Psychson) for example. To secure this szenario. You'd have to make sure that the Sticks are not infected by any kind of malware. Or use a USB condom. No OS will ever be able to secure itself from such attacks.