Dennis Fisher and Mike Mimoso discuss the post-RSA news, including the MySQL bug, the progress of the OpenSSL overhaul and the wildly entertaining House hearing on crypto backdoors.
Tag Archives: OpenSSL
OpenSSL Past, Present and Future
Heartbleed made the world notice what kind of shape OpenSSL development was in from a financial and resources standpoint. In the year since, the project has been funded enough to hire full-time engineers and a crucial refactoring of the codebase has the project in the right direction.
Heartbleed: One Year On
When news of the Heartbleed vulnerability broke this time last year, it was a watershed moment for the Internet and especially for security.
OpenSSL, the fundamental layer of encryption used by major websites around the world, was found to be flawed. Through a specific type of attack, a victim’s personal data including passwords, financial credentials could be stolen.
While the discovery of a vulnerability in OpenSSL didn’t come as much of a surprise to those who work in the security industry – after all, completely secure code is a rarity. Instead, the shock was the extent of the vulnerability, with around 60% of the entire web at risk.
Now, a year on, I’d love to be able to say that we’ve learned many lessons from Heartbleed and that the web is now a more secure place. Sadly, it’s not as simple as that.
Public awareness remains a major issue for Internet security. Recent research from password security developer Dashlane indicates that a year on, 86% of American’s have not heard of Heartbleed.
Dashlane spoke to AVG’s Chief Strategy Officer, Todd Simpson, about their results.
Video
The State of Online Security One-Year After Heartbleed
However, awareness is just one issue. Months after Heartbleed broke, I wrote of several further vulnerabilities in OpenSSL that had also emerged. Although each vulnerability discovered is theoretically a vulnerability fixed, it highlights the fact that this is still much work to be done. This is particularly true of open source software.
Open source software has several major benefits and will be around for a long time yet, but vulnerabilities such as Heartbleed demonstrate that there is risk and responsibility for all of us to protect the systems we have come to rely on.
Why has there been so little progress in securing OpenSSL and similar open source systems since Heartbleed appeared?
In my opinion, the issue lies within the very nature of open source software. OpenSSL is incredibly useful and has been adopted throughout the world, but how many people pay for OpenSSL, or donate time and money to keep it functional and secure? Not so many.
The OpenSSL Project does a great job finding and fixing vulnerabilities when they appear but in order to truly move the dial for Internet security, we need more investment.
Right now, the hands of the world’s online safety is in the hands of only a few coders working in small teams. That simply won’t do.
In April last year I wrote a blog highlighting a number of ways that we can all work together to improve the security of open source software.
Ultimately, it comes down to the fact that vulnerabilities will always exist; it’s up to all of us to take responsibility for our security.
The mysterious OpenSSL vulnerability has been patched
All users of OpenSSL 1.0.2 should upgrade immediately to version 1.0.2a. In the advisory published on their website the OpenSSL vulnerability is called “ClientHello sigalgs DoS (CVE-2015-0291)”. If a client connects to an OpenSSL 1.0.2 server and renegotiates with an invalid signature algorithms extension, a NULL pointer dereference will occur. This can be exploited in a DoS attack against the server.
According to OpenSSL’s Security Policy, a “high severity issue” includes issues affecting common configurations which are also likely to be exploitable. Examples include a server DoS (like this one), a significant leak of server memory (Heartbleed), and remote code execution.
OpenSSL promises that such issues “will be kept private and will trigger a new release of all supported versions”. They will attempt to keep the time these issues are private to a minimum, but the goal would be “no longer than a month” where this is something that can be controlled, and significantly quicker if there is a significant risk or we are aware the issue is being exploited.
The OpenSSL vulnerability has been reported on February 26th and the fix was released yesterday (March 19th), so well within the limit.
If this was no surprise, this advisory comes with something everyone was expecting: the FREAK vulnerability, which was initially categorized as “low severity”, has been reclassified as “high severity”. This was initially classified low because it was originally thought that servers with RSA export cipher suite support were rare: a client was only vulnerable to a MITM attack against a server which supports an RSA export cipher suite. Recent studies have shown that RSA export cipher suites support is far more common.
The patch comes also with fixes for a dozen or so vulnerabilities categorized as “moderate” and “low” severity.
Our recommendation is to update to version 1.0.2a immediately. Now that the vulnerability is public, it is to be expected that cybercriminals will try to exploit it.
The post The mysterious OpenSSL vulnerability has been patched appeared first on Avira Blog.
OpenSSL Mystery Patch is No Heartbleed
The anticipated high severity patch in OpenSSL is for a denial-of-service vulnerability in the recently released version 1.0.2 that can crash a client or server with a malformed certificate.
OpenSSL: Patch for secret “high severity†vulnerability
And indeed, in order to avoid being again in the news, the OpenSSL Foundation is set to release later this week several patches for OpenSSL, fixing undisclosed security vulnerabilities, including one that has been rated “high” severity.
Matt Caswell of the OpenSSL Project Team announced that OpenSSL versions 1.0.2a, 1.0.1m, 1.0.0r, and 0.9.8zf will be released Thursday.
“These releases will be made available on 19th March,” Caswell wrote. “They will fix a number of security defects. The highest severity defect fixed by these releases is classified as “high” severity.”
OpenSSL has been hit hard and the trust in it and in open source in general has been severely shaken in the last 12 months.
Last year in April, Heartbleed (CVE-2014-0160) was discovered in older versions of OpenSSL, but still highly used, which allowed hackers to read the sensitive contents of users’ encrypted data, such as financial transactions, instant messages and even steal SSL keys from Internet servers or client software that were running the affected versions of OpenSSL.
Two month later, in June the same year, a Man-in-the-Middle (MITM) vulnerability (CVE-2014-0224) was discovered and fixed. However, the vulnerability wasn’t quite as severe as the Heartbleed flaw, but serious enough to decrypt, read or manipulate the encrypted data.
In October last year, POODLE (CVE-2014-3566) (Padding Oracle On Downgraded Legacy Encryption) was discovered in the obsolete Secure Sockets Layer (SSL) v3.0 that could allow an attacker to decrypt contents of encrypted connections to websites. When exploited, it allows an attacker to perform a man-in-the-middle attack in order to decrypt HTTP cookies. The POODLE attack can force a connection to “fallback” to SSL 3.0, where it is then possible to steal cookies, which are meant to store personal data, website preferences or even passwords.
Just weeks ago, the latest vulnerability, FREAK (CVE-2015-0204) (Factoring Attack on RSA-EXPORT Keys) was discovered in the SSL protocol that allowed an attacker to force SSL clients, including OpenSSL, to downgrade to weaken ciphers that can be easily broken. Needless to say that such a weak encryption could potentially allow them to eavesdrop on encrypted networks by conducting man-in-the-middle attacks. This time, pretty much every big software vendor was affected: Apple, with its MacOS, iPhone and iPad, Google with Android and Chrome and last but not least, Microsoft with all versions of Windows.
Due to its widespread use, OpenSSL is considered an important software project and is ranked first under the Linux Foundation’s Core Infrastructure Initiative. Because of its complexity, high usage and lack of in-depth security reviews, companies like Google, Facebook and Cisco are heavily sponsoring this project in order to avoid being again affected by long forgotten bugs.
Well, for OpenSSL seems that this is starting to pay off.
The post OpenSSL: Patch for secret “high severity” vulnerability appeared first on Avira Blog.
OpenSSL Security Audit Ready to Start
NCC Group Cryptography Services announced it will shortly begin an audit of OpenSSL.
FREAK: All Windows versions are affected too
We wrote about the new SSL vulnerability called FREAK – Factoring RSA Export Keys – affects around 36% of all sites trusted by browsers and around 10% of the Alexa top one million domains, according to computer scientists at the University of Michigan.
Android, iOS and a lot of embedded devices that make use of the affected SSL clients (including Open) are in danger of having their connections to vulnerable websites intercepted.
The two most used operating systems for smartphones, tablets, laptops and embedded devices are in good company. Yesterday, Microsoft made known that all its supported Windows versions are also affected due to the presence of the vulnerability in the Windows Secure Channel (SChannel) – the Microsoft own implementation of SSL/TLS:
- Windows Server 2003
- Windows Vista
- Windows Server 2008
- Windows 7
- Windows 8 and 8.1
- Windows Server 2012
- Windows RT
Microsoft published an TechCenter an advisory where the problem is analyzed and solutions are offered. Also a patch is promised to fix all supported operating systems.
What does it mean for the user?
It means that if you are in Windows and make use of the vulnerable SSL libraries delivered by default, your connection to the affected servers can be intercepted. If you use Internet Explorer to visit www.freakattack.com you will be surprised to see this:
What should the users do?
We do not recommend messing up with the standard cryptography settings of Windows (or any operating systems) unless you know what you are doing (and there is a just hand full of people that actually do). You should try a browser that is not affected (like Chrome, which was updated in the meanwhile) and apply the patches for operating system and browsers that will come in the next few days.
The post FREAK: All Windows versions are affected too appeared first on Avira Blog.
Security experts are FREAKing out: new OpenSSL vulnerability
As any good and mind blowing (for most people) vulnerability, it has a nice name – FREAK, a CVE number – CVE-2015-0204 and a dedicated website https://freakattack.com/ .
FREAK – Factoring RSA Export Keys – affects around 36% of all sites trusted by browsers and around 10% of the Alexa top one million domains, according to computer scientists at the University of Michigan.
This time, the vulnerability can allow hackers to perform a Man In The Middle(MITM) attack on traffic routed between a device that uses the affected version of OpenSSL and many websites, by downgrading the encryption to an easy to crack 512 bits (64KB).
A connection is vulnerable if the server accepts RSA_EXPORT cipher suites and the client either offers an RSA_EXPORT suite or is using a version of OpenSSL that is vulnerable to CVE-2015-0204.
To be affected, devices must use the vulnerable version of OpenSSL. The problem is that OpenSSL is embedded sometimes in the firmware of the device like those running Apple’s iOS, Google’s Android. This makes the patching anything else than trivial. IfApple and Google will hurry up to patch their devices, not the same is going to happen with embedded devices that have the affected OpenSSL library in a firmware burned in a chip.
How is the attack happening?
If an attacker can monitor the traffic flowing between vulnerable devices (that is, running the vulnerable OpenSSL) and websites (that use the same vulnerable OpenSSL) they could inject code which forces both sides to use 512-bit encryption, which they can then crack in a matter of hours using the power of cloud computing.
It would then be technically pretty straightforward to launch a MITM by pretending to be the official website.
OpenSSL released a patch to the problem in January 2015, while Apple plans to do so next week and Google has released one to its Android partners.
As you can see, it is not trival to perform the MITM attack: special skills, a special environment and special tools are required to make use of this vulnerability. So, this makes FREAK a more theoretical vulnerability.But, this doesn’t mean that it is less dangerous.
However, as many times in the past, good intentions are badly implemented and the page freakattack.com is generously helping attackers to find which servers are affected. On that page the researchers from University of Michigan have published the top 10K domains listed by Alexa.com website.
Who is affected?
Websites that support RSA export cipher suites (e.g., TLS_RSA_EXPORT_WITH_DES40_CBC_SHA) are at risk to having HTTPS connections intercepted.
You can check whether a website supports RSA_EXPORT suites using the SSL FREAK Check available at this page.
The post Security experts are FREAKing out: new OpenSSL vulnerability appeared first on Avira Blog.
OpenSSL Fixes Eight Security Vulnerabilities
The OpenSSL Project has released several new versions of the software that fix eight security vulnerabilities, including several certificate issues and a couple of denial-of-service flaws. The patches included in OpenSSL 1.0.0p, 1.0.1k and 0.98zd are not for critical or high-risk vulnerabilities, but they do fix some interesting vulnerabilities. Two of the bugs are rated moderate and the other […]