iMessages is supposed to be bulletproof. Apple claims its end-to-end encrypted messaging service, as well as its video messaging app FaceTime, are so secure that even it lacks the means to decrypt the data sent between sender and receiver. No backdoors, no workarounds, nothing.
But security researcher QuarksLAB suggested on Thursday that Apple’s playing a kind of semantic game, that iMessages isn’t as seamlessly secure as claimed and that Apple could in fact lay hands on your encrypted data if it really, really wanted to. In short, say the researchers, while they’re not claiming Apple in fact reads iMessages, they “can … if they choose to, or if they are required to do so by a government order.”
In short — and the exhaustively analytical version is here — QuarksLAB says the problem’s by design: Apple controls the key infrastructure, thus “they can change a key anytime they want,” giving them access to your content. What’s more, Apple conveys the messages, so it has your metadata, too.
Apple’s response? “iMessage is not architected to allow Apple to read messages,” Apple spokesperson Trudy Muller told AllThingsD. “The research discussed theoretical vulnerabilities that would require Apple to re-engineer the iMessage system to exploit it, and Apple has no plans or intentions to do so.”
“Re-engineer” is a strong, fancy, quintessentially PR word. It implies doing something significant, like rebuilding the app from the ground up. But it could also mean doing something minor, like changing a line of code, or even a character in a line of code — anything at all, really. It’s a perfectly scalable term.
Regardless, it’s important to remember with these generalized, dichotomous claims about an app being either “secure” or “not secure,” that nothing is perfectly secure. It’s a question of probabilities: How secure is this mechanism for communication versus that one? How much do you trust a company like Apple not to “re-engineer” its product on the sly, at the behest of a power-player (like the NSA) in the interest of “national security”?
I use iMessages daily with impunity. I don’t worry about companies like Apple or Google poking around in my life because I’m boring. Maybe that’s naive. Probably it is. And regardless, privacy principles matter. But you’re taking inherent risks, however slight, if you use any messaging service, whatever the vendor’s claims about privacy guarantees. That’s the world we live in.
Update: QuarksLAB reached out to me to clarify that it doesn’t mean to suggest what it’s alleging Apple could theoretically do vis-a-vis iMessages is strictly a security issue, so much as a privacy issue.