At the beginning of the year, I did something I’ve never done before: I made a new year’s resolution. From here on out, I pledged, I would install only digitally signed software I could verify hadn’t been tampered with by someone sitting between me and the website that made it available for download.
It seemed like a modest undertaking, but in practice, it has already cost me a few hours of lost time. With practice, it’s no longer the productivity killer it was. Still, the experience left me smarting. In some cases, the extra time I spent verifying signatures did little or nothing to make me more secure. And too many times, the sites that took the time to provide digital signatures gave little guidance on how to use them. Even worse, in one case, subpar security practices of some software providers undercut the protection that’s supposed to be provided with digitally signed code. And in one extreme case, I installed the Adium instant messaging program with no assurance at all, effectively crossing my fingers that it hadn’t been maliciously modified by state-sponsored spies or criminally motivated hackers. More about those deficiencies later—let’s begin first with an explanation of why digital signatures are necessary and how to go about verifying them.
By now, most people are familiar with man-in-the-middle attacks. They’re waged by someone with the ability to monitor traffic passing between an end user and a website—for instance, a hacker sniffing an unsecured Wi-Fi connection or the National Security Agency sniffing the Internet backbone. When the data isn’t encrypted, the attacker can not only read private communications but also replace legitimate software normally available for download with maliciously modified software. If the attack is done correctly, the end user will have no idea what’s happening. Even when Web connections are encrypted with the HTTPS standard, highly skilled hackers still may be able to seed a website with malicious counterfeit downloads. That’s where digital signatures come in.