Christopher Soghoian writes: > I have asked Google's policy team, repeatedly, about what capabilities they > have for intercepting Hangout conversations, and I always get the same > vague no comment. > > Although Google is a clear transparency leader when it comes to reporting > aggregate stats on the # of requests that they receive, they still suck > when it comes to actually discussing their technical surveillance > capabilities, as well as the legal standards they follow when providing > surveillance assistance.
I sympathize with your frustration about Google and other companies' unwillingness to talk about their interception capabilities. In the particular case of Hangouts, it seems clear that the Hangout data is encrypted only between the user and Google, and not end-to-end. If so, intercepting Hangouts is even easier for Google than intercepting Skype calls is for Microsoft, since they don't even have to tamper with the key exchange process. They can just program their servers to passively record cleartext data already in their possession. It's disconcerting to see what a low priority secure end-to-end encryption continues to be for most designers of communications systems. (There might be technical reasons, too -- like wanting to transcode video, translate it, add captions, etc., but if people won't talk about the subject at all, we might never know the exact balance of factors that led to their decisions.) Two challenges for end-to-end encryption which have been discussed on this list are that many people want to access particular communications systems from multiple devices, and they may expect to use some services with a web browser instead of by installing a native client. The former means they might expect to access a service from a device where their private key isn't available (and, if they manage to copy the private key onto many devices, the risk of key compromise goes way up); the latter means that they're at risk of receiving a fresh backdoored version every time they connect. But we may be able to solve both of these things to some extent. A thornier challenge is that articulated demand for end-to-end crypto is very low, and arguably _falling_. So even though many of us have strongly criticized Skype's security model for years, they've felt no obvious embarrassment or need to change it, and others have felt no compunction about introducing new products with even lower levels of cryptographic protection, or even with explicit backdoors (like current work at ETSI on next-generation GSM voice encryption)! If Google _were_ willing to comment, they might say that very few users had voiced any objection to the Hangout security model, and that the product continues to be adopted on a huge scale, providing incremental security benefits relative to using a telephone. -- Seth Schoen <sch...@eff.org> Senior Staff Technologist https://www.eff.org/ Electronic Frontier Foundation https://www.eff.org/join 454 Shotwell Street, San Francisco, CA 94110 +1 415 436 9333 x107 -- Unsubscribe, change to digest, or change password at: https://mailman.stanford.edu/mailman/listinfo/liberationtech
