Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

State-run SSL certificate authorities make Congress nervous about web security

Ian Paul | June 11, 2015
Congress is losing sleep over the possibility other nations could endanger web security, and now it wants the four major browser makers to weigh in. The House of Representatives' Committee on Energy and Commerce recently sent letters to Apple, Google, Microsoft, and Mozilla with questions about how the backbone of HTTPS security could be violated.

Hackers could then use those fake certificates to create a man-in-the-middle (MITM) attack where users think they are connecting to Google or Bank of America but are actually handing their login details over to state-sponsored hackers.

Modern browsers have methods to detect potential MITM problems even with legitimate SSL certificates, but there's still a chance some users may be fooled and have their security compromised.

To defend against this possibility, Congress is asking the major browser makers whether government-owned CAs should be restricted in the kinds of certificates they issue. Instead of being allowed to issue a certificate for any site on the web, a government CA would only be allowed to issue certificates for its specific government domain. France, for example, could only issue SSL certificates for "gouv.fr"--"gouv.fr" is the French equivalent of ".gov" for American government sites.

Then if a browser saw a certificate for Twitter coming from a French government-owned CA, the browser could automatically reject that certificate as fraudulent.

Although Congress doesn't come out and say it, U.S. lawmakers are likely worried that nations like China, Iran, and Russia may try to carry out these kinds of attacks in their respective countries, as well as against U.S. interests.

Experts debate the consequences

While Congress ponders the problems with SSL certificates, security experts have also been debating the effectiveness of placing restrictions on CAs, including privately owned certificate issuers. "This is an idea that people have discussed for a long time," said Matthew D. Green, a cryptographer and research professor at Johns Hopkins University. "If it was implemented correctly it would certainly help to protect against some of the really, really bad compromises we've seen."

Kenneth White, security researcher and co-director of the Open Crypto Audit Project, isn't as convinced that a scheme like this could work. Nevertheless, he says, something needs to change. "The public CA trust system is already fundamentally broken according to many of us in the security world," White said. "I'm not sure if regulation is the right way versus the trust organizations themselves doing some sort of policing."

Currently, when fraudulent certificates start appearing online they are invalidated relatively quickly by all the major browser makers — though not always. In extreme cases, like the Diginotar hack, all certificates from that CA are blacklisted and the company may go out of business.

Although effective, the problem with blacklisting is that a fraudulent SSL certificate can still be in use for some time before anyone notices. Plus, says Green, browser makers are generally hesitant to blacklist an entire CA since it will break numerous websites around the world that people use every day.

 

Previous Page  1  2  3  Next Page 

Sign up for CIO Asia eNewsletters.