Security researchers have uncovered a newly discovered bug in Oracle's Java framework that allows attackers to bypass important security protections designed to prevent malware attacks.
The security improvements were introduced in Java 7 Update 10, and they came after a spate of in-the-wild attacks exploited fully patched versions of Java. Those allowed crooks to surreptitiously install malware on the computers of unsuspecting people using Java browser plugins. By default, the change required end users to manually allow the execution of Java code not digitally signed by a trusted authority. Users also had the ability to prevent any unsigned Java applet from running at all. Some security experts praised Oracle for adding the feature because it promised to drastically reduce the success of attacks that exploit security bugs in Java.
"Unfortunately, the above is only a theory," security researcher Adam Gowdiak wrote on Sunday, referring to the way the protections are supposed to block untrusted code from running on end-user computers. "In practice, it is possible to execute an unsigned (and malicious!) Java code without a prompt corresponding to security settings configured in Java Control Panel."