Microsoft password research has fatal flaw
I wrote yesterday about a report from Microsoft researchers, which goes against established password security best practices. The new guidance from the Microsoft researchers makes sense to me, because it fits how I handle password management already. However, at least one security expert feels that there is a fatal flaw that makes the new password advice impractical: You.
Almost every aspect of computer security and privacy seems to come back to that one fundamental issue. You—the user—are the weakest link in the security chain. No matter how effective a security process or tool has the potential to be, user error can undermine the whole thing and render the security useless.
In a nutshell, the Microsoft researchers assert that the default advice to use unique, complex passwords for every site and service you use doesn’t work. Users can’t remember that many complex passwords, so instead they opt to ignore the advice entirely and use the same often ridiculously simple password everywhere, increasing their exposure to risk and compromise. What the Microsoft researchers propose is that people group credentials based on their importance or access to sensitive data and feel free to re-use simple passwords for accounts that don’t really matter.
In this case, it’s not necessarily that the research is wrong. It’s just that the research is sadly too complex for the vast majority of users to implement effectively.
“The problem remains that many non-tech savvy people don't know [on] which sites to use their 'standard' password, and which sites to use their super-secure password,” commented one person on my Google+ page.
Tim Erlin, Tripwire’s director of IT security and risk strategy, agrees. “The conclusions and analysis of this research is fundamentally flawed because it relies on an assumption that users are capable of determining which accounts are ‘of value.’ This is clearly demonstrated by the tendency of many users sharing sensitive data on Facebook.”
Erlin explained that a user might reasonably conclude that an email account doesn’t contain “sensitive data,” and therefore doesn’t warrant a more secure password under the password management system proposed by the Microsoft researchers. However, an attacker with access to an email account can execute a password reset for other accounts like banks or financial institutions. Failing to use a strong password for a seemingly innocuous email account leaves you open to severe compromise—just ask Mat Honan.
As I stated at the top, the advice from the Microsoft researchers makes sense to me because it’s how I manage passwords already. It seems like every website requires me to register, and most of them I couldn’t care less about so I use the same, simple password for all of those. For my financial accounts, email, and social networks, though, I use much stronger, unique passwords—thereby limiting the volume of complex passwords I have to keep track of.
Put another way, the same naiveté and cavalier attitude that leads so many users to reject the established password security best practices will result in this new approach failing for them as well. Any solution that expects users to fully understand the risks, assign value to different accounts, and then follow through with thorough execution of the solution is set-up for failure.