Security researchers at Kaspersky Lab have discovered a new piece of malware that uses a PNG (Portable Network Graphics) image to hide malicious code.
The malware is distributed by email in a clean PDF, which includes a link to a .zip file that contains the malicious image, along with other files. This type of attack is not new, but is similar to previously observed attacks where cybercriminals distribute .exe or .zip files that contain a .pdf extension in the filename.
While the attack technique is not new, the delivery method is, revealing that bad actors are doing their best to find new ways to ensure their malicious code goes undetected by anti-virus products.
Last year, information-stealing malware Stegoloader, which caused havoc among North American healthcare organizations, was also observed using a PNG image file to hide its main module. The image file was downloaded using a hardcoded URL in the initial payload and was automatically decompressed and decrypted to execute the malware and infect the compromised computer.
In this new scenario, Kaspersky’s Thiago Marques explains that attackers have moved beyond simple phishing attacks and attempted to hide the malicious payload in encrypted files that have a known file format, to avoid raising user suspicion.
Kaspersky researchers discovered that the PDF file was generated from HTML content and even managed to find the URL of the tool used. They also observed that the malicious link in the file prompts users to download a JAR file, which downloads a ZIP file containing other files.
While analyzing the content of the .zip file, the researchers noticed that three of the files didn’t have an extension, but included a PNG file header, which shows the file type that will be used in order to open it. The file was a solid color image of only 63 x 48 pixels, but had a large file size of 1.33 MB.
The researchers identified the function that loads the PNG files to the memory and discovered that it is also responsible for decrypting and executing the extracted binary the RunPE technique. Also known as VBInject or VBCrypt, it ensures that the malicious code is executed in the context of another process, in this case iexplore.exe, while also making it difficult to detect the nefarious operation.
Further analysis revealed that the PNG image was only 179 bytes, and that the remaining content in the file was the encrypted malicious code.
The malicious payload hosted in the PNG image cannot be executed without its launcher, meaning that it cannot be used as the main infector, Marques says. With the malware usually delivered to the victim’s mailbox, it has to be installed by a different module, but the technique is efficient at hiding malicious code inside images to avoid automated detection and to make analysis more difficult.
Kaspersky Lab researchers also note that five different files were associated with this attack, namely Trojan.Win32.KillAv.ovo, HEUR:Trojan.Win32.Generic, Trojan-Downloader.Win32.Banload.cxmj, Trojan-Downloader.Win32.Agent.hgpf, and HEUR:Trojan-Downloader.Java.Generic.
The appearance of this threat also reveals that Brazilian attacks are constantly evolving and that they are becoming more complex and efficient, the researcher notes. Just as Proofpoint revealed in a recent report, attackers are viewing people as the best exploit, and users should take extra care when opening emails or attachments received from unknown sources.
In a December 2015 SecurityWeek column, Bill Sweeney, US financial services evangelist of BAE Systems Applied Intelligence, explained that social engineering techniques have become more personalized, technologically advanced and ultimately successful. He also presented a series of effective counter-measures that enterprises can employ to improve their security.
Related: Lossy Image Compression Can Hide Malicious Code in PDF Files: Researcher
Related: Cybercriminals Use SVG Files to Distribute Ransomware
