LIVE · cybersecurity feed
Live wire
Acronis warns of actively exploited flaw in its cPanel backup pluginOracle September 2026 Critical Security Patch Update addresses 672 CVEsCVE-2026-76461 · U.S. CISA adds Cisco Secure Email Gateway flaw to its Known Exploited Vulnerabilities catalogHackers target WordPress sites via third-party WooCommerce pluginCVE-2026-51990 · Hackers exploit Tencent app flaw to deploy GrayRabbit malwareCVE-2026-42016 · CISA Adds 5 Exploited Flaws in Artifactory, ScreenConnect, RouterOS to KEVCVE-2026-85102 · Dutch NCSC: Critical Check Point VPN flaws exploitation is imminentAI Enables Mass Generation of Personalized Fraud EmailsCVE-2026-85706 · Critical GitLab Vulnerability Exploited in Internet-Wide ProbesCVE-2025-66516 · Metasploit Wrap Up: This One Goes to Sixteen!
vulnerability

DeepZero: Open-source hunting for vulnerable Windows drivers

DeepZero is an open-source engine that automates the search for exploitable Windows kernel drivers. You point it at a folder of binaries and it parses them, pulls them apart, scans them, throws most of them away, and asks a language model whether what survives can be attacked. Pipelines are written in YAML, the code is Python 3.11 and up. Rehman Ahmadzai, who maintains the project, said DeepZero h

zeroday.news ·

DeepZero, an open-source engine designed to automate the discovery of exploitable vulnerabilities in Windows kernel drivers, has identified multiple confirmed flaws within a subset of the Snappy Driver Installer corpus. Some of these findings are reportedly still undergoing the disclosure process. The project, maintained by Rehman Ahmadzai, is available for free on GitHub.

The engine's bundled pipeline specifically targets "bring your own vulnerable driver" (BYOVD) scenarios, where an attacker loads a legitimately signed driver containing a flaw to gain kernel-level access. DeepZero is built as a target-agnostic pipeline orchestrator, meaning its underlying framework is not limited to Windows kernel drivers or a single architecture, although the included research pipeline is tailored for them.

The vulnerability detection process involves a seven-stage pipeline. Initially, it parses Portable Executable (PE) headers. The second stage filters for kernel-mode drivers that expose Input/Output Control (IOCTL) surfaces, which are the control codes user space sends to a driver. Drivers already cataloged on loldrivers.io, a public list of known vulnerable drivers, are then excluded.

Following these initial filters, the Ghidra reverse engineering platform runs headless to decompile the remaining binaries. Semgrep rules then scan the exported C source code. A "pick_top_10" reduction step trims the field to a top candidate tier before a language model assesses the exploitability of the surviving binaries. The AI evaluation is placed at the end to leverage the context gathered by earlier stages, such as binary decompilation and the exclusion of known vulnerable or unsupported binaries.

Ahmadzai noted that confirming a finding can sometimes require physical hardware. Many drivers create their device object only when their associated hardware is enumerated. Without the specific device on a machine or virtual machine, the driver's exposed functionalities may be unreachable, making it appear benign from a report alone.

To address this, the decompile stage records whether the `IoCreateDevice` function is called and if it's invoked by `DriverEntry` or a function that `DriverEntry` calls. A "true" value indicates the device will appear on any machine loading the driver, allowing for straightforward confirmation of a finding. A "false" value typically suggests a plug-and-play callback creates the device, necessitating the specific hardware for confirmation. If `IoCreateDevice` is not found, neither value is recorded.

vulnerabilityai
ShareXLinkedInWhatsAppFacebook

More News

view all →
vulnerability

The vulnpocalypse rains iBugs down on Apple with record-setting number of patches

September Patch Tuesday part 2?

vulnerabilityhigh

Acronis warns of actively exploited flaw in its cPanel backup plugin

Acronis disclosed a high-severity Linux local privilege escalation vulnerability in its backup plugin for cPanel, WebHost Manager (WHM), and Plesk that may be exploited in the wild. [...]

vulnerabilitycritical

Oracle September 2026 Critical Security Patch Update addresses 672 CVEs

Oracle addresses 672 CVEs in its September 2026 Critical Security Patch Update with 673 patches, including 104 critical updates. Key Takeaways The September 2026 Critical Security Patch Update (CSPU) contains fixes for 672 unique CVEs in 673 security updates 104 issues (15.5% of all patches) were assigned a critical severity rating Oracle E-Business Suite received the highest number of patches at

CVE-2026-76461critical

U.S. CISA adds Cisco Secure Email Gateway flaw to its Known Exploited Vulnerabilities catalog

U.S. Cybersecurity and Infrastructure Security Agency (CISA) adds Cisco Secure Email Gateway flaw to its Known Exploited Vulnerabilities catalog. The U.S. Cybersecurity and Infrastructure Security Agency (CISA) added a Cisco Secure Email Gateway flaw, tracked as CVE-2026-76461 (CVSS score of 9,8), to its Known Exploited Vulnerabilities (KEV) catalog. Cisco disclosed a critical zero-day CVE-2026-76

patch

Mythos has made 2026 patching hell. It might make 2027 a breeze

Gartner sees huge amounts of technical debt paid down, and better scanning that could make software safer sooner

breach

The modern attack chain: Rethinking Google Workspace security in the age of AI

Over the past two months, I’ve written about the Vercel breach and the Composio breach separately. Both offer lessons to learn on their own. But reading them together, I keep coming back to the same observation: these aren’t isolated incidents. They’re the same attack, run twice, against different targets, where email was not the entry point into the workspace. And once you see the pattern clearly