PoC Released for DirtyDecrypt Linux Kernel Vulnerability

Summary

Proof-of-concept code has been released for DirtyDecrypt (DirtyCBC), a Linux kernel vulnerability discovered by the V12 security team that allows local attackers to escalate privileges to root. The flaw is a missing copy-on-write (COW) guard in the rxgk_decrypt_skb component of the RxGK subsystem used by the Andrew File System (AFS), likely corresponding to CVE-2026-31635 with a CVSS score of 7.5, which was patched in April. The vulnerability only affects distributions with CONFIG_RXGK enabled, such as Arch Linux, Fedora, and openSUSE, though exploitation in container environments could enable pod escape attacks.

Full text

Proof-of-concept (PoC) code is now available for another Linux kernel vulnerability that could allow attackers to elevate their privileges to root. Dubbed DirtyDecrypt (aka DirtyCBC), the exploit comes from the V12 security team, which discovered it earlier this month, after fixes were rolled out in April. The V12 team has not shared a CVE identifier for the security defect, but noted that it is a missing copy-on-write (COW) guard in the rxgk_decrypt_skb component of the RxGK subsystem. RxGK is a security class for the RxRPC network protocol used by the Andrew File System (AFS) and OpenAFS, which relies on the GSSAPI framework to provide authentication, confidentiality, and integrity protection. Due to the missing COW guard, oversized response authenticators are accepted, which results in data being written to the memory of privileged processes or to the page cache of privileged files, such as SUID binaries, Moselwal notes. As Tharros Labs senior principal vulnerability analyst Will Dormann points out, the underlying issue could be CVE-2026-31635 (CVSS score of 7.5), a Linux kernel vulnerability disclosed on April 24, when patches were rolled out for mainline Linux builds.Advertisement. Scroll to continue reading. DirtyDecrypt only affects distributions that have CONFIG_RXGK compiled in and enabled, such as Arch Linux, Fedora, and openSUSE. In container platforms, all worker nodes running a vulnerable distribution could provide attackers with a path to escape the pod, Moselwal says. According to V12, the flaw is a variant of the recently identified CopyFail, DirtyFrag, and Fragnesia Linux kernel bugs, all of which grant root access on vulnerable systems. Disclosed last week and officially tracked as CVE-2026-46300, Fragnesia affects the XFRM ESP-in-TCP subsystem. It allows attackers to overwrite sensitive system files and gain root privileges. The Dirty Frag exploit published earlier this month chains two vulnerabilities in the Linux kernel, including one that affects the RxRPC component, to elevate privileges to root. Copy Fail, which was disclosed in late April, enables an attacker to modify the in-memory copies of setuid-root binaries, providing root shell access. Threat actors started exploiting it shortly after disclosure. Related: Researcher Drops MiniPlasma Windows Exploit for Unpatched 2020 CVE Related: Exploitation of Critical NGINX Vulnerability Begins Related: Google Detects First AI-Generated Zero-Day Exploit Related: Over 40,000 Servers Compromised in Ongoing cPanel Exploitation Written By Ionut Arghire Ionut Arghire is an international correspondent for SecurityWeek. Daily Briefing Newsletter Subscribe to the SecurityWeek Email Briefing for the latest cybersecurity threats, trends, and expert insights. More from Ionut Arghire Admins of Bulletproof Hosting Service Used by Russian Hackers Arrested in Netherlands266,000 Affected by Data Breach at Radiology Associates of RichmondLaravel-Lang Packages Poisoned for Malware DeliveryDocketWise Data Breach Impacts 143,000Over 5,500 GitHub Repositories Infected in ‘Megalodon’ Supply Chain Attack‘Underminr’ Vulnerability Lets Attackers Hide Malicious Connections Behind Trusted DomainsGrafana Says Codebase and Other Data Stolen via TanStack Supply Chain AttackCisco Patches Critical Vulnerability in Secure Workload Latest News AppOmni’s Marlin AI Brings Autonomous Investigation to SaaS SecurityIranian APT Targets Aviation, Software Companies With Updated Tools185,000 Likely Impacted by 7-Eleven Data BreachAnthropic Expands Claude’s Enterprise Security Governance With 28 New IntegrationsHackers Exploited KnowledgeDeliver Zero-Day for Web Shell DeploymentWatch on Demand: Threat Detection & Incident Response Summit – All Sessions AvailableOpen Source DockSec Uses AI to Cut Through Vulnerability Noise in Docker ImagesLithuania Suspects Foreign Involvement in Data Leak of Over 600,000 National Register Entries Trending Daily Briefing NewsletterSubscribe to the SecurityWeek Email Briefing to stay informed on the latest threats, trends, and technology, along with insightful columns from industry experts. Virtual Event: Threat Detection and Incident Response Summit On-Demand Delve into big-picture strategies to reduce attack surfaces, improve patch management, conduct post-incident forensics, and tools and tricks needed in a modern organization. Register Webinar: Third-Party Risk in Practice June 4, 2026 Organizations are investing heavily in third-party risk management, but breaches, delays, and blind spots continue to persist. Join this live webinar as we examine the gap between how organizations think their third-party risk programs are performing and what’s actually happening in practice. Register People on the MoveJoe Chen has become Chief Technology Officer at Trellix.Usercentrics has named Pawan Hegde as COO and Elena Ignatova as CPTO.SecureAuth has named Mark van Oppen as Chief Revenue Officer.More People On The MoveExpert Insights Caught Off Guard: Securing AI After It Hits Production As enterprises rush AI projects into production, security teams are increasingly being forced into reactive mode. (Joshua Goldfarb) Cyber Resilience is the New Business Continuity Plan The organizations best prepared to face disruption are those that align security, continuity and risk management around what the business cannot afford to lose. (Steve Durbin) Enhancing Data Center Security Without Sacrificing Performance For AI data centers, where the stakes are the highest and performance constraints are the tightest, security and performance are no longer a zero-sum game. (Nadir Izrael) Is the SOC Obsolete, and We Just Haven’t Admitted It Yet? Many AI-first enterprises have already embraced sovereign architectures for general AI initiatives; cybersecurity—and the SOC—should be next. (Danelle Au) The Mythos Moment: Enterprises Must Fight Agents with Agents Only with the right platform and an agentic, AI-driven defense, will enterprises be able to protect themselves in the agentic era. (Etay Maor) Flipboard Reddit Whatsapp Whatsapp Email

Indicators of Compromise

• cve — CVE-2026-31635
• cve — CVE-2026-46300

Entities