No news is bad news for two-factor logins

Assume SecurID is broken?

It's been a week since RSA dropped a vaguely worded bombshell on 30,000 customers that the soundness of the SecurID system they used to secure their corporate and governmental networks was compromised after hackers stole confidential information concerning the two-factor authentication product.

For seven days, reporters, researchers, and customers have called on RSA, and its parent corporation EMC, to specify what data was lifted – or at the very least to say if it included details that could allow government or corporate spies to predict the one-time passwords that SecurID tokens generate every 60 seconds. And for seven days, the company has resolutely refused to answer. Instead, RSA has parroted Security 101 how-tos about strong passwords, support-desk best practices, and the dangers of clicking on email attachments.

Officials from RSA and EMC have steadfastly refused to give yes or no answers to two questions that have profound consequences for the 40 million or so accounts that are protected by SecurID: Were the individual seed values used to generate a new pseudo-random number exposed and, similarly, was the mechanism that maps a token's serial number to its seed leaked?

Without the answers to those two basic questions, RSA customers can't make educated decisions about whether to continue relying on SecurID to prevent unauthorized logins to their sensitive networks. After all, if the breach on RSA's servers exposed the seeds and the mapping mechanism, SecurID customers have lost one of the factors offered by the two-factor authentication product.

An RSA spokesman released an updated statement earlier this week that said in part: “Our investigation to date has revealed that the attack resulted in certain information being extracted from RSA’s systems. Even with this information being extracted, RSA SecurID technology continues to be an effective authentication solution for customers.” (Notice the statement didn't say “an effective two-factor authentication solution.”)

The statement went on to say that revealing additional details “could enable others to try to compromise our customers’ RSA SecurID implementations, so we are not disclosing further information.”

Translation: Yes, we were hacked, and yes, the hackers made off with confidential information that compromises the security of a product you've spent huge amounts of money on, but you'll just have to trust us that you're still safe.

In the wake of this information blackout, the prudent thing for customers to do is to assume that SecurID seeds have been lifted, and to also assume that the mechanism that maps a particular token's serial number to its individual seed has also been taken. That means if attackers can trick individual SecurID users into giving out the number printed on the back of their token, its two-factor protection has been broken. The same applies if a company's database of serial numbers is breached.

That assumption would be consistent with an advisory RSA sent to customers on Monday urging them to strengthen the personal identification numbers that are used along with a user ID and the one-time password, since the PIN would be the single factor of authentication left.

SecurID's two-factor authentication may not be broken, but until RSA comes clean and provides some yes or no answers to two simple questions, it's better to assume it is. The network security you preserve may be your own.