
Cloud Vulnerability DB
A community-led vulnerabilities database
CVE-2026-21684 is an undefined behavior vulnerability in the CIccTagSpectralViewingConditions() function of the iccDEV library, a set of libraries and tools for interacting with International Color Consortium (ICC) color management profiles. It affects iccDEV versions prior to 2.3.1.2 and was discovered by researcher xsscx, reported on November 21, 2025, with a fix merged on November 24, 2025, and formally disclosed on January 7, 2026. The vulnerability carries a CVSS v3.1 base score of 7.1 (High) (GitHub Advisory, Feedly).
The root cause is a reliance on undefined behavior (CWE-758) combined with improper input validation (CWE-20) in the CIccTagSpectralViewingConditions() copy constructor within IccProfLib/IccUtil.cpp at line 555. Specifically, the code performs a floating-point-to-integer conversion where the floating-point value is -nan (Not a Number), which is outside the representable range of int, triggering undefined behavior as detected by UndefinedBehaviorSanitizer (UBSan). An attacker can craft a malicious ICC color profile file that, when processed by an application using the iccDEV library, triggers this undefined behavior path. The fix involved correcting the copy constructors for CIccTagSpectralViewingConditions in pull request #225 (GitHub Issue, GitHub PR).
Successful exploitation requires user interaction — a user or application must process a maliciously crafted ICC color profile. The primary impact is high availability loss (denial of service via application crash or undefined behavior) and low integrity impact, with no confidentiality impact. Applications and services that automatically process user-supplied or externally sourced ICC profiles are at greatest risk, as the undefined behavior can lead to unpredictable program states beyond simple crashes (GitHub Advisory, Feedly).
A proof-of-concept demonstrating the issue is publicly available via GitHub Issue #216, which includes a reproducible test case using the Cat8Lab-D65_2degMeta.icc profile and the iccRoundTrip tool built with AddressSanitizer and UBSan (GitHub Issue). There is no evidence of in-the-wild exploitation at this time, and the vulnerability is not listed in the CISA Known Exploited Vulnerabilities (KEV) catalog. The EPSS score is approximately 0.055% (0.000550), indicating a low probability of exploitation in the near term (Feedly).
Cat8Lab-D65_2degMeta.icc as a basis)..icc file to a target user or upload it to a service that automatically processes ICC profiles using the iccDEV library (versions prior to 2.3.1.2).CIccTagSpectralViewingConditions() constructor.IccProfLib/IccUtil.cpp:555 triggers undefined behavior, potentially causing a crash (denial of service) or unpredictable program state (GitHub Issue, GitHub Advisory)..icc profile files in directories processed by applications using iccDEV; specifically profiles with anomalous spectral viewing conditions tags.IccProfLib/IccUtil.cpp:555; UBSan runtime error messages such as runtime error: -nan is outside the range of representable values of type 'int'.iccRoundTrip, iccApplyProfiles, or similar iccDEV-based utilities when processing user-supplied profiles (GitHub Issue).Upgrade iccDEV to version 2.3.1.2 or later, which contains the fix for the undefined behavior in CIccTagSpectralViewingConditions() (PR #225, merged November 24, 2025). No workarounds are available for unpatched versions. Organizations should prioritize patching systems that process untrusted or externally sourced ICC color profiles, and consider validating ICC profile inputs at the application layer as a defense-in-depth measure (GitHub Advisory, GitHub PR).
The vulnerability received limited but focused attention from the security community, with automated CVE tracking services (CIRCL, VulDB, Vulners, CVEFeed) picking up the disclosure shortly after publication. A CISA vulnerability bulletin (SB26-012) for the week of January 5, 2026 referenced the CVE. Social media activity was minimal, with a Bluesky post from an automated CVE tracking account noting the disclosure (Feedly).
Source: This report was generated using AI
Free Vulnerability Assessment
Evaluate your cloud security practices across 9 security domains to benchmark your risk level and identify gaps in your defenses.
Get a personalized demo
"Best User Experience I have ever seen, provides full visibility to cloud workloads."
"Wiz provides a single pane of glass to see what is going on in our cloud environments."
"We know that if Wiz identifies something as critical, it actually is."