Hacking your own antivirus for fun and profit (Safe browsing gone wrong)
Bullguard has a safe browsing feature to prevent their users from entering websites that contain malware, phishing or other malicious content. The idea in itself is great, but earlier this year I made discovery about, how this feature can be abused for malicious purposes. I have made multiple attempts at responsible disclosure since I discovered this vulnerability, but it still hasn’t been fixed, so after 4 months of attempted responsible disclosure, I have decided to make full disclosure of the vulnerability
In other words: The vulnerability allows an attacker to perform reflected Cross Site Scripting attacks on Google, Yahoo and Bing’s domains, if the victim uses bullguard. Imagine the following scenario:
1: An attacker inserts a malicious vector to perform session hijacking though a domain that is vulnerable to XSS in a GET parameter
2: The attacker manages to get enough clicks and abuse the SEO of the vulnerable domain to get his attack vector to show up as a result on google, yahoo and bing
or alternatively: The vector gets so much SEO that it executes by itself from certain legitimate keywords!
4: The attacker has performed a succesful session hijacking attack against your google, microsoft or yahoo account!
XSS can be used for much more, but that’s another topic… Google it, if you’re curious.
Obviously, I decided to report the vulnerability to bullguard themselves, but unfortunately, the only available way of contacting them is though their customer service… Better than nothing, and worth a try though! I made a clear description about the issue and asked the customer service to pass the information through to their developers, or other relevant co-workers in their company.
After no patching had been done, I decided to give it a try again 3 months later, and it turned out that the development team had never responded to the customer service, who notified them about the issue. The customer service gave it another try, but nothing happened
16/05/2018: Vulnerability reported for the first time
02/07/2018 and 15/08/2018: Got in touch with bullguard to tell them that they were still vulnerable
14/09/2018: Public disclosure
Protip: Who says this vulnerability only affects bullguard? If your antivirus has a similar feature, try it out and remember to report the vulnerability responsibly to the vendor, if it’s vulnerable