A surprising 35 percent of websites around the world are still using SHA-1 though barely 45 days remain before some of the major browsers stop trusting certificates signed with the hash algorithm altogether.
The estimate, from cybersecurity vendor Venafi, is based on an analysis of data from some 11 million publicly visible websites.
The analysis showed that while most of the popular websites have migrated away from SHA-1, a troublingly large number has not. Based on recent numbers from Netcraft showing over 173 million active websites on the Internet, Venafi extrapolated its results to mean there are about 61 million sites that rely on the deprecated certificates.
Such sites can expect transactions and traffic to be disrupted in multiple ways starting January 1, 2017. Google, for instance, has said that its Chrome browser will drop all support for SHA-1 certificates starting that date. Chrome will block all certificates signed with the algorithm triggering a fatal network error, Google has previously warned.
Like Mozilla and the others, Google stopped treating SHA-1 certificates as secure a long time ago, and has been discontinuing support for it in a phased manner. Starting earlier this year, Chrome has displayed a certificate error warning for sites using SHA-1 certificates that were issued on or after January 1, 2016. That is the date before which all certificate authorities were supposed to stop issuing SHA-1 certificates.
Through the year, Google has been ramping up those warnings with a view to weaning websites off SHA-1 completely by the end of this year.
Others have been following a similar approach. Starting with Windows 10 Anniversary Update, Microsoft’s Edge and Internet Explorer browsers have been removing the address bar lock icon on sites relying on SHA-1. Though the sites have continued to work, the browsers no longer consider them to be secure. That will change further in Feb 2017 when both browsers will block SHA-1-signed certificates altogether.
The moves by the browser makers are a response to long-standing concerns over SHA-1 security. Noted security researchers such as cryptography expert Bruce Schneier have warned about SHA-1 being vulnerable to collision attacks — where two messages generate the same hash value — as far back as 2005.
Last year, researchers from the national research institute for math and computer science in the Netherlands, France’s Inria, and the Nayang Technology University in Singapore, showed how a full-scale collision attack on SHA-1 could be carried out for about $100,000 using cloud based hardware. The research showed that attacks against SHA-1 were not only possible sooner than most experts had anticipated, but also could be performed at much lower cost as well.
The transition deadline for SHA-1 is long overdue, says Kevin Bocek, vice president of security strategy and threat intelligence at Venafi. The National Institute of Standards and Technology called for the elimination of SHA-1 in 2006, over security concerns, he notes.
“Successful attacks on SHA-1 are well within reach of nation states and other sophisticated adversaries, and these allow them to ‘mint’ trusted SHA-1 certificates,” Bocek says.
As one example of the damage that can be caused by such certificate counterfeiting, Bocek points to a 2012 incident where attackers were able to widely distribute the Flame malware sample using forged Microsoft MD5 certificates.
The key to getting rid of SHA-1 certificates is figuring out how and where they are being used, he says. Large organizations are especially likely to have little visibility over the extent to which such certificates are being used, and need to have a plan for discovering and eliminating them, he says.
Jai Vijayan is a seasoned technology reporter with over 20 years of experience in IT trade journalism. He was most recently a Senior Editor at Computerworld, where he covered information security and data privacy issues for the publication. Over the course of his 20-year ... View Full Bio