Skip to main content
Emerging ThreatsMalware & Ransomware

CISA Warns of Exploited Flaws in Ransomware, Cryptomining Attacks

Rows of computer servers and equipment racks in a brightly-lit server room.

CVE-2026-90436, CVE-2026-90817, CVE-2026-40217 and CVE-2024-10441—four actively exploited vulnerabilities—were added this week to CISA’s Known Exploited Vulnerabilities (KEV) catalog, each tied to ongoing ransomware, cryptomining, or LLMjacking activity.

CVE-2026-90436: Apache Axis 1.x exploitation and post-compromise activity

CISA added CVE-2026-90436 (CVSS 9.8) to the KEV list after reports that attackers are exploiting the unsafe reflection vulnerability in Axis 1.x to execute arbitrary code. Rapid7 documented an incident flagged on November 13, 2026, in which attackers leveraged the Axis flaw against an Atlassian Confluence Data Center instance that used the Axis library as part of its XML‑RPC and SOAP APIs.

Atlassian noted that, to be vulnerable, a Confluence instance must have the "Remote API (XML‑RPC & SOAP)" enabled, and that the setting "was turned off by default for new installations starting in Confluence 3.0" and that the feature "was fully removed in Confluence 10.0." Because Axis 1.x has reached end‑of‑life, CISA’s advisory recommends migration to a different SOAP engine such as Apache Axis 2/Java or using a build of Axis with the July 12, 2026 commit patch applied.

In Rapid7’s account, the flaw was weaponized to run shell commands and scripts fetched from temporary Pinggy and Cloudflare Tunnel domains, followed quickly by environment enumeration. "Just over three minutes after the first observed execution, the attackers were authenticating to MySQL with Confluence's own database credentials, querying page bodies, application links, and attachments with names associated with credentials, tokens, VPN configurations, and private keys," Rapid7 said. Rapid7 added that "the attacker's scripts, likely AI‑generated, then targeted files in Confluence's default location for uploaded page attachments (/var/atlassian/application-data/confluence/attachments/v4/), parsed Office documents, searched for credentials, and hunted for payment card numbers, using Luhn checks to filter out false positives."

Rapid7 also reported that attackers mapped internal services, queried Atlassian Crowd for users and groups using a Confluence application credential, attempted Kerberos AS‑REP Roasting (unsuccessfully), acquired a service account credential, queried a production cloud data warehouse, and—about 12 hours after first exploitation—signed in to Microsoft 365 as an integration service account using legacy authentication from "192.238.253[.]176."

CVE-2026-90817: REDCap code injection, transient web shells, and cryptomining

CVE-2026-90817 (CVSS 9.8) affects Vanderbilt University’s REDCap and is a code injection flaw that can allow unauthenticated remote code execution. CISA described the exploitation vector: "An attacker could exploit this via crafted HTTP requests that access an unintended controller route from a public survey context and supply a crafted file-path or stream parameter during data import processing." CISA further noted that "Exploitation requires knowledge of a valid public survey hash."

Safe Security reported in a late‑October analysis that an unknown threat actor abused this flaw to "stage a transient web shell, followed by an in‑memory loader that retrieved XMRig disguised as image files, masqueraded the miner as syslog‑ng, and established persistence through cron and an added SSH key." The activity ties the REDCap flaw directly to cryptomining campaigns that establish on‑host persistence.

CVE-2026-40217: LiteLLM unprotected channel and LLMjacking campaigns

CVE-2026-40217 (CVSS 8.8) is an unprotected alternate channel vulnerability in BerriAI LiteLLM. The flaw allows remote execution via bytecode rewriting at the /guardrails/test_custom_code URI. Google Threat Intelligence Group reported that threat actors are leveraging this weakness to elevate limited inference keys into full administrative credentials and to establish persistence as part of LLMjacking attacks targeting organizations across financial services, retail, transportation, and technology in North America, South America, and the Asia Pacific region.

CVE-2024-10441: Synology DSM, ransomware, and C3RB3R

CVE-2024-10441 (CVSS 9.8) is an improper encoding or escaping of output flaw in Synology DiskStation Manager (DSM) that Synology disclosed earlier in November 2026 as being exploited in ransomware attacks against NAS devices. Those incidents have been found to result in deployment of C3RB3R ransomware, according to CISA’s KEV advisory.

What this means for technologists, policymakers, and affected enterprises

  • Technologists and security teams: CISA set federal mitigation deadlines—apply fixes or necessary mitigations for DSM, REDCap, and Axis flaws by November 21, 2026, and for the LiteLLM flaw by December 2, 2026—underscoring urgency for patching, configuration hardening (for example, disabling remote XML‑RPC/SOAP APIs where present), and detection of the specific post‑exploit behaviors Rapid7 and Safe Security observed.
  • Procurement and operations leaders at affected enterprises: The Axis 1.x recommendation to migrate to a supported SOAP engine or apply the July 12, 2026 commit patch highlights a nontrivial remediation decision where vendors or legacy platforms include end‑of‑life components that cannot be patched conventionally.
  • Policy and risk officials: The diversity of outcomes—cryptomining through REDCap, LLMjacking via LiteLLM, and ransomware on Synology NAS—illustrates that a single KEV update can cover multiple threat objectives (cryptomining, data exfiltration, and destructive encryption), and that deadlines for federal agencies are being used to compress remediation windows.

The KEV additions make plain that exploitation is not theoretical: defenders are seeing ephemeral web shells, AI‑assisted post‑exploit scripts, credential harvesting, and the reuse of legacy remote APIs to gain footholds. With federal deadlines set and active campaigns already tied to each vulnerability, the immediate questions are operational—who will patch or mitigate in time, and how quickly will defenders detect the telltale post‑exploit behaviors that Rapid7 and others have documented?

Original story