CISA Adds Synology, LiteLLM, Axis and REDCap Flaws to KEV Catalog

CISA added four exploited flaws in Synology DSM, LiteLLM, Apache Axis 1.x and Vanderbilt REDCap to KEV, tied to ransomware, cryptomining and LLMjacking attacks.
Table of Contents
    Add a header to begin generating the table of contents

    CISA added four actively exploited vulnerabilities to its Known Exploited Vulnerabilities catalog on November 18, covering Synology DSM, BerriAI LiteLLM, Apache Axis 1.x and Vanderbilt REDCap. The flaws are linked to ransomware on network storage devices, cryptomining and attacks that steal cloud AI credentials.

    Four Exploited Flaws Span Storage, AI Gateways and Research Software

    The four entries are CVE-2024-10441 in Synology DSM, CVE-2026-40217 in LiteLLM, CVE-2026-90436 in Apache Axis 1.x and CVE-2026-90817 in REDCap. Public proof-of-concept code exists for the LiteLLM, Axis and REDCap flaws. CISA ordered forensic triage under BOD 26-04 for the Synology, Axis and REDCap entries.

    CVE-2024-10441 and C3RB3R Ransomware on Synology NAS Devices

    CVE-2024-10441 carries a CVSS score of 9.8 and sits in the system plugin daemon of Synology DSM and BeeStation OS, where it allows unauthenticated code execution. Synology disclosed earlier this month that attackers had used it to deploy C3RB3R ransomware on NAS devices. Fixed builds are DSM 7.2.2-72806-1, DSM 7.2.1-69057-6 and BSM 1.1-65374.

    CVE-2026-40217 and LLMjacking Through the LiteLLM Guardrails Endpoint

    CVE-2026-40217 (CVSS 8.8) is a code injection flaw in BerriAI LiteLLM reachable through the /guardrails/test_custom_code endpoint, where an AST sandbox bypass allows arbitrary code to run. Google Threat Intelligence Group reported attackers using it to convert limited inference keys into administrator credentials in LLMjacking attacks. Victims included financial services, retail, transportation and technology firms in North America, South America and Asia Pacific. LiteLLM builds after April 8, 2026 fix the flaw, and Red Hat has issued errata.

    Cryptominers Reach Axis 1.x and REDCap Servers

    Two of the four flaws were used to install cryptocurrency miners, in different campaigns and against different software.

    CVE-2026-90436 and the End-of-Life Apache Axis 1.x Problem

    CVE-2026-90436 (CVSS 9.8) is an unsafe reflection flaw in Apache Axis 1.x. Attackers exploited it alongside the 4EverAble2Shell flaws in N-able N-central to deploy cryptominers. Axis 1 is end-of-life and unpatched, so the options are migrating to Axis2/Java or applying the July 12 commit.

    CVE-2026-90817 and the XMRig Loader Disguised as Image Files

    CVE-2026-90817 (CVSS 9.8) is a code injection flaw in Vanderbilt REDCap that is reachable with a valid public survey hash. Safe Security observed an actor stage a web shell and an in-memory loader that delivered the XMRig miner disguised as image files. The malware masqueraded as syslog-ng and persisted through cron jobs and SSH keys. Safe Security first reported REDCap exploitation in late October. Fixed versions are 16.0.49, 17.3.10 and 17.4.4.

    CISA KEV Federal Deadlines of November 21 and December 2

    Federal civilian agencies have until November 21 to remediate the Synology, REDCap and Axis entries, and until December 2 for LiteLLM. The deadlines are short, particularly for the Synology, Axis and REDCap flaws, which share a date only days after the listing.

    The mix of targets in a single KEV update shows how varied the exploitation picture is. Ransomware operators went after NAS devices, cryptomining actors went after an unmaintained Java library and a research data platform, and credential thieves targeted AI infrastructure. REDCap’s exposure carries particular weight because the platform holds research and clinical data.

    The BOD 26-04 triage requirement for the Synology, Axis and REDCap entries goes beyond patching: agencies must look for evidence that systems were already compromised, which fits the fact that all three flaws had been exploited before CISA’s listing. The LiteLLM flaw’s later December 2 deadline gives agencies more time, though Google Threat Intelligence Group’s report shows attackers were already using it against firms on three continents.

    Organizations outside the federal government are not bound by the deadlines, but KEV listings are widely used as a prioritization signal. The LiteLLM entry in particular reflects the growing attention to AI gateways as a source of high-value credentials, since a single inference key can grant access to paid model services.

    Related Posts