CVE-2026-46520: ImageMagick Out-of-Bounds Heap Write Denial of Service

Technical summary

CVE-2026-46520 is an out-of-bounds heap write vulnerability (CWE-122, CWE-787) in ImageMagick's multi-image processing logic. When the library reads multiple images with differing dimensions without proper bounds validation, a write operation can exceed allocated heap memory. The flaw is network-exploitable with no authentication or user interaction required, affecting both the legacy 6.x and current 7.x release lines. The vulnerability was patched in versions 6.9.13-48 and 7.1.2-23.

Business impact

Organizations deploying ImageMagick in production pipelines—such as web applications processing user-uploaded images, media conversion services, or batch image processing workflows—face availability risk. An attacker can send malformed image files to trigger denial of service, disrupting image upload features, thumbnail generation, or automated media workflows. In sandboxed or memory-constrained environments, heap corruption may enable local privilege escalation or code execution, though the primary impact vector is availability.

Affected systems

ImageMagick versions prior to 6.9.13-48 (6.x branch) and 7.1.2-23 (7.x branch) are vulnerable. Any application or service that links against or calls ImageMagick to process untrusted image files is at risk, including web servers, content management systems, image resizing services, document processors, and media automation tools. Both direct library integration and server-side image processing APIs are in scope.

Exploitability

The vulnerability is exploitable over the network without authentication. An attacker needs only to craft and submit image files with mismatched dimensions to a service that processes them with vulnerable ImageMagick versions. No user interaction is required; automated batch processing increases exposure. However, exploitation complexity is low because it relies on triggering a common operation (multi-image reading) rather than a complex state or configuration.

Remediation

Upgrade ImageMagick to version 6.9.13-48 or later for the legacy branch, or version 7.1.2-23 or later for the current branch. Verify with vendor advisories before deployment. Interim mitigation includes disabling multi-image processing, validating image dimensions before processing, implementing request size limits, and running ImageMagick in a restricted or sandboxed environment with minimal privileges.

Patch guidance

Priority should be placed on upgrading to patched versions 6.9.13-48 (6.x) or 7.1.2-23 (7.x). Organizations using the legacy 6.x branch should plan a migration roadmap to 7.x if extended support is needed. Test patches in non-production environments first, as ImageMagick updates may affect image processing behavior or output quality. If custom builds or distributions are used, verify that your vendor or package maintainer has backported the fix.

Detection guidance

Monitor for image processing errors, memory faults, or application crashes when handling multi-image files or batched image uploads. Implement logging around ImageMagick calls to detect unusual or malformed input. Network-based detection can identify attempts to upload images with inconsistent or suspicious metadata (e.g., dimension mismatches). Intrusion detection systems may flag exploitation attempts if they include known image fuzzing payloads. Heap corruption detection tools and address sanitizers are useful in development and staging environments.

Why prioritize this

This vulnerability carries a CVSS 3.1 score of 7.5 (HIGH severity) due to network exploitability, low attack complexity, and high availability impact. Although not yet listed in the CISA Known Exploited Vulnerabilities catalog, the straightforward exploitation method (sending malformed images) and widespread deployment of ImageMagick in web services make it attractive to attackers. Organizations processing untrusted image uploads should prioritize patching within 30 days.

Risk score, explained

The score reflects a network-accessible denial of service vector with no barriers to exploitation. The absence of confidentiality or integrity impact (C:N, I:N) keeps the score from critical range, but high availability impact (A:H) and universal applicability (S:U) justify HIGH severity. The attack vector (AV:N), low complexity (AC:L), and lack of privilege requirements (PR:N, UI:N) indicate broad exploitability in real-world deployments.

Frequently asked questions