Loading vulnerability…
CVE-2026-81983 is an out-of-bounds write vulnerability in the PDF image-processing pipeline of Adobe Acrobat and Acrobat Reader on Windows and macOS. A 32-bit integer overflow in calculations for 16-bit image samples allows an undersized image-data stream to pass validation. A subsequent byte-swapping loop retains an excessively large sample count and accesses memory beyond the allocated buffer. Opening a specially crafted PDF triggers the vulnerable rendering path and can cause memory corruption or arbitrary code execution with the current user's privileges. Affected releases include Acrobat and Acrobat Reader Continuous 26.002.21900 and earlier, and Acrobat 2024 Classic 24.001.30383 and earlier. The vulnerability has a CVSS 3.1 score of 7.8.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
No public exploits tracked yet. Mallory keeps watching.
No public exploit code observed for this vulnerability.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
19 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
An out-of-bounds write vulnerability affecting Adobe Acrobat 2024 versions 24.001.30383 and earlier, and Acrobat and Acrobat Reader versions earlier than 26.002.21901. Adobe released fixes in Acrobat 2024 version 24.001.30429 and Acrobat/Acrobat Reader version 26.002.21901. The advisory recommends updating to these or later versions; it does not describe exploitation or specific security impacts.
An integer overflow in the Acrobat.dll image-processing pipeline causes an out-of-bounds write when a victim opens a crafted PDF containing a 16-bit image. A 32-bit byte-size calculation wraps around, allowing an undersized image stream to pass validation while a byte-swap loop processes an excessively large sample count. This corrupts memory and could enable code execution with the victim’s privileges. The reported CVSS score is 7.8, and exploitation requires user interaction. Adobe fixed the issue in bulletin APSB26-141 on September 8, 2026: Acrobat and Reader Continuous 26.002.21900 and earlier were patched in 26.002.21901, and Acrobat 2024 Classic 24.001.30383 and earlier were patched in 24.001.30429.
An integer overflow in the Adobe Acrobat and Acrobat Reader image-processing pipeline causes an out-of-bounds write when a victim opens a crafted PDF. A 32-bit image-size calculation wraps, allowing an undersized data stream to pass validation before a byte-swap loop overruns a buffer. Exploitation could execute code with the victim’s privileges. The article reports a CVSS score of 7.8 and an Adobe critical severity rating. Adobe fixed the flaw using 64-bit calculations and rejecting sample counts above INT_MAX. Security bulletin APSB26-141, released September 8, 2026, provides fixed Continuous version 26.002.21901 and Acrobat 2024 Classic version 24.001.30429.
An integer overflow vulnerability in Adobe Acrobat/Reader that causes memory corruption when a malicious PDF is opened. The post describes it as a critical zero-day discovered through a week of AI-assisted fuzzing, but provides no details about exploitation in the wild or remediation.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.