Cybercriminals commonly known as Magecart hackers are distributing malicious PHP web shells masqueraded as a favicon to inject JavaScript skimmers into online shopping platforms running Magento 1.x in order to steal financial information from their users.

According to Malwarebytes security researcher Jérôme Segura, the web shells in question are tracked as Smilodon or Megalodon and are used to dynamically load JavaScript skimming code via server-side requests into online stores. This technique allows a threat actor to evade detection.

This latest campaign involves a PHP-based web shell disguised as a favicon (“Magento.png”), with the malware inserted into compromised sites by replacing the legitimate shortcut icon tags with a path to the fake PNG file. This web shell then retrieves the next-stage payload (a credit card skimmer) from an external host. The credit card skimmer used in this campaign is similar to another version dubbed ‘Cardbleed’ used in an attack in September 2020.

Malwarebytes linked this latest campaign to Magecart Group 12 based on methods, techniques, and procedures observed in the attacks. The researchers also noticed that the domain name (zolo[.]pw) involved in the campaign is “hosted on the same IP address (217.12.204[.]185) as recaptcha-in[.]pw and google-statik[.]pw, domains previously associated with Magecart Group 12.”

“There are a number of ways to load skimming code but the most common one is by calling an external JavaScript resource. When a customer visits an online store, their browser will make a request to a domain hosting the skimmer. Although criminals will constantly expand on their infrastructure it is relatively easy to block these skimmers using a domain/IP database approach,” Malwarebytes wrote.

“In comparison, the skimmer we showed in this blog dynamically injects code into the merchant site. The request to the malicious domain hosting the skimming code is not made client-side but server-side instead. As such a database blocking approach would not work here unless all compromised stores were blacklisted, which is a catch-22 situation. A more effective, but also more complex and prone to false positives approach, is to inspect the DOM in real time and detect when malicious code has been loaded.”