37 ms·

I thought the idea was to get data out/transmit out of an "infected" computer when it has no public network access: and not actually sniffing normal traffic. so

I thought the idea was to get data out/transmit out of an "infected" computer when it has no public network access: and not actually sniffing normal traffic. so it assumes a compromised air-gapped computer, is that not a concern of infosec? is that a purely theoretical scenario that won't happen ? why?

marcan_42 5y ago
If you're at the level where that is a concern, you should be putting your air-gapped computers inside a bunker. This should be obvious to anyone doing threat modeling at this level. It is clear that there are a myriad ways of modulating data in environmental emissions out of a computer; we don't need a new paper from this guy for every single light, RF, sound, and heat emitting component in a computer. We know. I can come up with as many of these as you want. Last time I talked about this I came up with a dozen or so, and after checking he'd done them all, except one: the camera LED. So I wrote a 5-minute tweet-sized PoC to modulate data using the camera LED. That would've been on par with his papers. If he actually did real science e.g. trying to estimate maximum theoretical data transfer rates with these techniques then his research would be valuable, but he isn't. He's just doing minimal proofs of concept. We all know these things work. Tempest for Eliza has been around for 20 years now and was a cooler demo than anything this guy has put out [1]. [1] http://www.erikyyy.de/tempest/ http://www.erikyyy.de/tempest/