A weakness has been reported in Check Point Firewall-1 NG with SmartDefense, which allows malware to bypass detection.

The weakness is caused due to a lack of RFC2397 support. This can be exploited to bypass the malware detection by sending malicious image files, which are base64 encoded and embedded in an HTML file according to the standard specified in RFC2397, which is supported by a number of client applications capable of rendering HTML files (e.g. email clients and browsers).

A PoC has been published, which embeds an image that attempts to exploit the GDI+ JPEG parsing vulnerability in Microsoft Windows.

NOTE: Content inspection software can generally be bypassed in many ways by obfuscating data and exploit code. However, this advisory describes lack of compliance with a widely deployed standard for embedding pictures in HTML files.

This has been reported to affect Check Point Firewall-1 NG R55 HFA08 with SmartDefense 541041226. Other versions may also be vulnerable.

Solution
The vendor recommends using the newly added option to block encoded images: "Enable Block Encoded images". Note: This may impact the functionality of some websites and emails.
Further details available in Customer Area

Provided and/or discovered by
Darren Bounds, Intrusense.

Changelog
Further details available in Customer Area

Original Advisory
http://www.intrusense.com/av-bypass/image-bypass-advisory.txt

Other references
Further details available in Customer Area

Deep Links
Links available in Customer Area