10 Facts: Secure Java For Business Use

Is Java safe to use? That's the refrain heard after every round of new zero-day vulnerabilities that get spotted in Java, followed days or weeks later by related patches from Oracle.

But the question still stands: Is the Java programming language -- which encompasses client-side desktop applications and Web browser extensions, embedded platforms, as well as Java running on smartphones such as Android -- safe to use? Or is it an over-targeted time bomb that's best avoided by anyone with an ounce of security sense?

Here are 10 related facts:

1. Security Concern: Client-Side Java

To be clear, the current Java security worries center on client-side Java, and the prevalence with which attackers have been finding and exploiting vulnerabilities in Java browser extensions. The latest threat has been the two zero-day vulnerabilities in Java 7 first publicly detailed last week, which allow attackers to run arbitrary code on vulnerable machines. Oracle Sunday released an update, dubbed Java 7 Update 11, that fixes or works around the flaws.

Monday, however, security firm Immunity reported that the fix from Oracle only repairs one of the two zero-day flaws. "Only one of the two bugs were fixed, making Java still vulnerable to one of the bugs used in the exploit found in the wild," said security researcher Esteban Guillardoy at Immunity in a blog post.

2. Second Zero-Day Vulnerability Remains A Vulnerability

But the other component wasn't patched per se, but rather addressed via new, default security settings in Java, which now require a user to authorize any Java applet that wants to run.

[ The attacks just keep on coming. Read Red October Espionage Network Rivals Flame. ]

Unfortunately, that "fix" now puts more security onus on users. "In theory, this should reduce the impact of malicious applets. However, because users can still expressly authorize these malicious applets, users may still be affected," said Jonathan Leopando, a technical communications specialist at Trend Micro, in a blog post.

Furthermore, the unpatched vulnerability remains. Using that bug, "an attacker with enough knowledge of the Java code base and the help of another zero day bug to replace the one [that's been] fixed can easily continue compromising users," said Guillardoy, provided the attacker launches the exploit using a signed Java applet.

3. DHS Recommends: Disable Java

In the wake of the Sunday part-patch, the Department of Homeland Security Monday said that from a risk standpoint, Java remains too hot to handle. "Unless it is absolutely necessary to run Java in Web browsers, disable it ... even after updating to 7u11," according to the DHS advisory (which also details exactly how to disable Java). "This will help mitigate other Java vulnerabilities that may be discovered in the future."

The tail end of the advisory encapsulates many security experts' current thinking: Disabling Java today ensures businesses won't be unknowingly compromised by future zero-day Java vulnerabilities. Or as Bogdan Botezatu, a senior e-threat analyst at security software vendor Bitdefender, put it via email: "As [Java] attacks are highly likely to hit from the Web, the absence of the plug-in would dramatically cut down on the attack surface."

4. Danger: Java Continues To Be Attack Magnet

Attack surface is the operative phrase, because zero-day Java vulnerabilities continue to be sought after by online criminals or anyone else seeking to exploit targeted PCs. "These types of vulnerabilities are attractive to criminals because Java is somewhat platform agnostic -- so the same vulnerability can be used to reliably exploit a variety of targets -- and historically, Oracle has been slow to release fixes, which maximizes the timeframe in which the exploit can be utilized," said Joe DeMesy, a senior analyst at information security consultancy Stach & Liu, via email.

Indeed, the Red October espionage malware (nicknamed "Rocra") first publicly detailed Monday by Kaspersky Lab includes an attack module for exploiting a Java vulnerability (CVE-2011-3544) that was patched in October 2011. But the most recent Rocra attack module designed to exploit the vulnerability was compiled in February 2012, reported security firm Seculert. That lag highlights how even after a patch had been released, attackers still expected to find exploitable machines four months later.