Researchers exposed a new cryptographic Secure Sockets Layer/Transport Layer Security (SSL/TLS) vulnerability, dubbed Logjam, but some security experts are downplaying the risks.
According to a new report, authored by researchers from Inria, Microsoft Research, Johns Hopkins University, University of Michigan and the University of Pennsylvania, the Logjam flaw is similar to the FREAK (Factoring attack on RSA-EXPORT Keys) vulnerability, first reported in March. As with FREAK, with Logjam, an attacker can potentially downgrade a TLS connection to a weaker export grade of cryptography that can be exploited.
"The attack is reminiscent of the FREAK attack, but is due to a flaw in the TLS protocol, rather than an implementation vulnerability, and attacks a Diffie-Hellman key exchange, rather than an RSA key exchange," the Logjam report stated.
The researchers commented that, based on their own analysis of published National Security Agency (NSA) leaks, it is likely the spy agency may have used a Logjam type of attack.
From a patching perspective, though Logjam is only now coming to public light, Microsoft patched it on May 12 with security bulletin MS15-055.
"Customers who apply the update, or have automatic updates enabled, will be protected," a Microsoft spokesperson told eWEEK. "We encourage all customers to apply the update to help stay protected."
Google fixed the issue with the Chrome 42 update, which debuted on April 15.
"We disabled TLS False-Start with Diffie-Hellman (DHE) in Chrome 42, which has been the stable version for many weeks now," Google engineer Adam Langley wrote in a mailing list message. "This will make their attack on vulnerable servers slightly harder."
Mozilla's patch for Firefox isn't out yet, but "we expect it to be published in the next few days," Richard Barnes, cryptographic engineering manager at Mozilla, told eWEEK.
Most of the coordination across browser vendors for Logjam was done, thanks to the researchers who found the bug, Barnes said.
"They provided valuable measurement data to the various browser vendors, which allowed us all to calibrate our response," Barnes said.
Security experts eWEEK contacted were not surprised by the Logjam disclosure.
Over the years, SSL Labs and SSL Pulse, Qualys services that provide tests and reports on the state of SSL/TLS deployment, have been tracking the issue that Logjam exposed, explained Ivan Ristic, director of application security research at Qualys.
"The use of obsolete cryptography—in this case, export suites and suites with DH parameters below 1,024 bits—is still common," Ristic told eWEEK.
SSL Labs started reporting on the low bit cryptography several years ago, warning that most sites use less than the desired 2,048 bits. Ristic noted, though, that anyone following good configuration practices from the last several years would have avoided this problem.
"We didn't make any changes to SSL Labs for this last attack, but all servers vulnerable to it would be getting an F because we were already detecting insecure crypo," Ristic said.
Chris Eng, vice president of research at Veracode, does not consider the likelihood of a Logjam attack to be high. End users need not worry too much, as long as they continue to run updated browsers, he said.
"The combination of factors that must coincide to enable a successful attack is unlikely in practice," Eng told eWEEK. "Server operators simply need to disable export-grade cipher suites, which has been a recommended practice for well over a decade."
Many companies have already removed the export-grade cipher suites in response to FREAK, Eng said. As with FREAK, the most important outcome of the Logjam research is that it will be another nail in the coffin for long-outdated encryption methods on the Internet, he added.
Tod Beardsley, security research manager at Rapid 7, commented that Logjam is a man-in-the-middle (MiTM) attack that is conceptually easier to execute then the POODLE vulnerability in SSLv3 that was disclosed in October 2014.
"It still relies on intentionally weak export-grade ciphers from that dark time in the early days of the Internet, so it's another good reminder of why those were a bad idea," Beardsley said.
Fundamentally, what Logjam represents is a breach of trust, Beardsley said. "Cryptography makes certain promises about confidentiality, and this vulnerability violates that trust," he said. "When a browser shows the user a happy green lock icon, we should be able to trust that nobody, not even state actors, can eavesdrop on that connection."
Sean Michael Kerner is a senior editor at eWEEK and InternetNews.com. Follow him on Twitter @TechJournalist.