Accidental Leak Reveals Chinese Hackers Have IE Zero Day
Google researcher's new fuzzer finds vulnerabilities in all browsers
January 3, 2011
A renowned Google researcher who this week released a new free fuzzer that so far has found around 100 vulnerabilities in all browsers says Chinese hackers appear to have gotten their hands on one of the same bugs he discovered with the tool.
Google's Michal Zalewski unleashed the so-called cross_fuzz tool on New Year's Day and announced the fuzzer to date uncovered more than 100 vulnerabilities, many of them exploitable, in all browsers.
In a bizarre twist, Zalewski says an accidental leak of the address of the fuzzer prior to its release helped reveal some unexpected intelligence, namely that "third parties in China" apparently also know about an unpatched and exploitable bug he found in IE with the fuzzer. It all started when one of cross_fuzz's developers, who was working on crashes in the open-source WebKit browser engine used in Chrome and Safari, inadvertently leaked the address of the fuzzer in one of the crash traces that was uploaded. That made the fuzzer's directory, as well as the IE test results from the fuzzer indexed by GoogleBot, he says.
Zalewski says he was able to confirm afterward that there were no downloads or discoveries of the tool. But on Dec. 30, he says, an IP address in China queried keywords included in one of the indexed cross_fuzz files, specifically two DLL functions, BreakAASpecial and BreakCircularMemoryReferences, associated with and unique to the zero-day IE flaw he found with the fuzzer.
"The person had no apparent knowledge of cross_fuzz itself, poked around the directory for a while, and downloaded all the accessible files; suggesting this not being an agent one of the notified vendors, but also being a security-minded visitor," Zalewski explained in his blog post. "The pattern is very strongly indicative of an independent discovery of the same fault condition in MSIE by unrelated means; other explanations for this pair of consecutive searches seem extremely unlikely."
Microsoft, meanwhile, said in a statement that now that information about the vulnerability is public, "the risk has now been amplified," but that it hasn't seen any signs of attack thus far. "Working with software vendors to address potential vulnerabilities in their products before details are made public reduces the overall risk to customers. In this case, risk has now been amplified. We will continue to investigate this issue and take appropriate action to help protect customers," said Jerry Bryant, group manager for response communications at Microsoft, in a statement. "Microsoft is investigating this potentially exploitable vulnerability and will take the appropriate steps to help protect customers. As always, we are closely monitoring the threat landscape and are not aware of any attempts to try and exploit the issue."
Anup Ghosh, founder and chief scientist at Invincea, says Zalewski's fuzzer appears to be sophisticated such that it can explore more of the state space of the browser document object model than a simple fuzzer can.
He says it's not really surprising that browsers can't handle unexpected input well. "If they do not handle unexpected input -- and most exploits fall into that category -- they can be susceptible to exploits that grant privileges to code that shouldn't have them, such as that from malicious websites," Ghosh says. "Time will tell how many of these bugs will become exploitable vulnerabilities. With this fuzz-testing tool now available, bug finders will have a new sophisticated tool at their disposal to help find zero-days."
The sheer complexity of a browser basically guarantees it will contain bugs, he says.
Zalewski says Microsoft had asked him to hold off on releasing the tool -- which he first alerted the company about in July -- but he went forward with his plan to release it in early January. "Vendor has acknowledged receiving the report in July (case 10205jr), but has not contacted me again until my final ping in December. Following that contact attempt, they were able to quickly reproduce multiple exploitable crashes, and asked for the release of this tool to be postponed indefinitely. Since they have not provided a compelling explanation as to why these issues could not have been investigated earlier, I refused," he blogged.
Zalewski says Microsoft was concerned with the PR ramifications of the fuzzer and its findings, and that it at first was unable to perform the same browser crashes he had reported to them with his tool. Microsoft in late December was able to find the same flaws, however, the researchers there told him, Zalewski says.
Have a comment on this story? Please click "Discuss" below. If you'd like to contact Dark Reading's editors directly, send us a message.
About the Author
You May Also Like
Unleashing AI to Assess Cyber Security Risk
Nov 12, 2024Securing Tomorrow, Today: How to Navigate Zero Trust
Nov 13, 2024The State of Attack Surface Management (ASM), Featuring Forrester
Nov 15, 2024Applying the Principle of Least Privilege to the Cloud
Nov 18, 2024The Right Way to Use Artificial Intelligence and Machine Learning in Incident Response
Nov 20, 2024