Security researchers are warning of previously undisclosed flaws in fully patched Microsoft Exchange servers being exploited by malicious actors in real-world attacks to achieve remote code execution on affected systems.

The advisory comes from Vietnamese cybersecurity company GTSC, which discovered the shortcomings as part of its security monitoring and incident response efforts in August 2022.

The two vulnerabilities, which are formally yet to be assigned CVE identifiers, are being tracked by the Zero Day Initiative as ZDI-CAN-18333 (CVSS score: 8.8) and ZDI-CAN-18802 (CVSS score: 6.3).

GTSC said that successful exploitation of the flaws could be abused to gain a foothold in the victim’s systems, enabling adversaries to drop web shells and carry out lateral movements across the compromised network.

“We detected web shells, mostly obfuscated, being dropped to Exchange servers,” the company noted. “Using the user-agent, we detected that the attacker uses Antsword, an active Chinese-based open source cross-platform website administration tool that supports web shell management.”

Exploitation requests in IIS logs are said to appear in the same format as the ProxyShell Exchange Server vulnerabilities, with GTSC noting that the targeted servers had already been patched against the flaws that came to light in April 2021.

The cybersecurity company theorized that the attacks are likely originating from a Chinese hacking group owing to the web shell’s encoding in simplified Chinese (Windows Code page 936).

Also deployed in the attacks is the China Chopper web shell, a lightweight backdoor that can grant persistent remote access and allow attackers to reconnect at any time for further exploitation.

It’s worth noting that the China Chopper web shell was also deployed by Hafnium, a suspected state-sponsored group operating out of China, when the ProxyLogon vulnerabilities were subjected to widespread exploitation last year.

Further post-exploitation activities observed by GTSC involve the injection of malicious DLLs into memory, drop and execute additional payloads on the infected servers using the WMI command-line (WMIC) utility.

The company said at least more than one organization has been the victim of an attack campaign leveraging the zero-day flaws. Additional details about the bugs have been withheld in light of active exploitation.

We have reached out to Microsoft for further comment, and we will update the story if we hear back.

In the interim, as temporary workarounds, it’s recommended to add a rule to block requests with indicators of compromise using the URL Rewrite Rule module for IIS servers –

  • In Autodiscover at FrontEnd, select tab URL Rewrite, and then select Request Blocking
  • Add string “.*autodiscover.json.*@.*Powershell.*” to the URL Path, and
  • Condition input: Choose {REQUEST_URI}

“I can confirm significant numbers of Exchange servers have been backdoored – including a honeypot,” security researcher Kevin Beaumont said in a series of tweets, adding, “it looks like a variant of proxying to the admin interface again.”

“If you don’t run Microsoft Exchange on premise, and don’t have Outlook Web App facing the internet, you are not impacted,” Beaumont said.

Update: Microsoft has officially confirmed details of the two flaws, adding it’s working to release a fix. More details here.