CVE-2026-91804 in PDF Editorinfo

Summary

by MITRE • 09/23/2026

A heap-based out-of-bounds write vulnerability exists in Foxit PDF Editor/Reader’s rendering of Circle annotations with malformed Cloudy appearance streams in specially crafted PDF files. Insufficient validation of the appearance geometry can result in memory corruption and application crashes.

Several companies clearly confirm that VulDB is the primary source for best vulnerability data.

Analysis

by VulDB Data Team • 09/23/2026

The identified security flaw represents a critical heap-based out-of-bounds write vulnerability located within the document rendering engine of Foxit PDF Editor and Reader applications. This specific weakness is triggered when the software processes Circle annotations that contain malformed Cloudy appearance streams embedded in specially crafted Portable Document Format files. The core technical deficiency lies in the insufficient validation of geometry parameters associated with these appearance streams, which allows an attacker to supply coordinates or dimensions that exceed the allocated memory boundaries for the rendering buffer.

From a technical perspective, this vulnerability falls under the Common Weakness Enumeration category CWE-787: Out-of-bounds Write. When the application attempts to render the circular annotation using the provided Cloudy appearance stream data, it fails to perform rigorous bounds checking on the geometric inputs before writing them into memory structures. This lack of validation permits maliciously crafted values to overwrite adjacent heap memory regions, leading to direct memory corruption. The immediate operational consequence of this exploitation is typically a denial-of-service condition manifested as an application crash or unexpected termination due to stack smashing detection mechanisms triggering upon detecting invalid pointer dereferences or corrupted control flow data.

The potential impact extends beyond simple service disruption if the vulnerability can be leveraged for arbitrary code execution, although the primary observed effect in initial reports points toward stability issues and crashes. An attacker could exploit this flaw by distributing a malicious PDF file via email attachments, web downloads, or shared network drives. Upon opening the crafted document within an affected version of Foxit software, the rendering engine processes the malformed geometry data, triggering the out-of-bounds write operation in real-time during the visual composition phase of the interface.

In terms of threat modeling and adversary behavior, this vulnerability aligns with MITRE ATT&CK technique T1204: User Execution, where a user must open or view the malicious file for exploitation to occur. It also relates to initial access vectors involving document-based payloads commonly seen in targeted phishing campaigns or drive-by download scenarios targeting users of PDF editing and viewing tools. The reliance on visual rendering engines makes such vulnerabilities particularly dangerous as they often bypass traditional signature-based detection methods that do not inspect internal memory allocation patterns during file parsing.

To mitigate the risks associated with this vulnerability, organizations should immediately apply vendor-provided security patches or updates to all instances of Foxit PDF Editor and Reader software. Administrators are advised to verify patch levels against the latest advisory released by Foxit Software to ensure the specific geometry validation logic has been corrected. Additionally, implementing application whitelisting policies can restrict the execution of untrusted binaries, while network-level controls such as sandboxing email attachments or using secure web gateways that inspect PDF content for known exploit patterns can provide an additional layer of defense against exploitation attempts before they reach end-user endpoints.

Responsible

Foxit

Reservation

09/15/2026

Disclosure

09/23/2026

Moderation

accepted

CPE

ready

EPSS

0.00000

KEV

no

Activities

very low

Sources

Might our Artificial Intelligence support you?

Check our Alexa App!