A newly disclosed flaw in the way OpenAI, Anthropic, and Google carried hidden AI reasoning between API calls let researchers recover internal reasoning and secrets from session logs, including API keys and passwords. The weakness affected encrypted reasoning objects used by the providers' reasoning APIs, where a block created in one session could be replayed into another and, during testing,
Enterprise defenses are tuned to catch the attacks that make noise. This year's data shows attackers winning by making none. According to Picus Labs' new Blue Report 2026, which measured more than 338 million real attack simulations across actual client production environments in the first half of 2026, defenses are having one of their strongest years yet. Average prevention effectiveness
Signal has introduced Automatic Key Verification, a new security feature that gives users a new way to ensure their encrypted chats haven't been intercepted. [...]
Adobe has shipped updates to address multiple critical security vulnerabilities impacting ColdFusion, Commerce, and Campaign Classic that, if successfully exploited, could result in arbitrary code execution and privilege escalation. The most severe of the flaws are listed below - CVE-2026-48362 (CVSS score: 10.0) - An operating system command injection vulnerability in ColdFusion that could
Public vulnerability advisory and supporting evidence for CVE-2026-67846 affecting the BOOM v3/v4 NBDTLB implementation.
Joomla Extension - tabaoca.org - Improper ACL implementation allows file operations in Cotton Cloud < 2.0.3 - Authenticated users could perform various file-related operations (read, delete, overwrite, re-assign permissions) on files owned by other users.
When Microsoft Excel imports a CSV file, it executes cells beginning with certain characters as formulas, giving such CSV files arbitrary execution. Velociraptor fails to sanitize such cells when exporting to CSV from various places such as the GUI, offline collector or data exports. It is not clear if the vulnerability is actually in Microsoft Excel treating a CSV data file as executable content, or if Velociraptor should be sanitizing the data to prevent Excel from executing it. However, si
The hunt_delete() VQL function allows deleting hunts. Velociraptor misapplied the permission check requiring only COLLECT_CLIENT (usually assigned to the "investigator" role) instead of the "DELETE_RESULTS" permission (usually only assigned to "administrators").
A rogue Velociraptor client can upload a malformed sparse file such that if the GUI attempts to expand the file, a panic occurs which may crash the server process. The problem is a Divide by Zero bug in the ShouldPadFile() function.
Nightmare Eclipse has released a new Microsoft Defender zero-day exploit named "ShieldBreak" after Microsoft released the August 2026 Patch Tuesday security updates. [...]
Joomla Extension - tabaoca.org - Improper ACL implementation allows file operations in Cotton Cloud < 2.0.2 - Unauthenticated users could perform various file-related operations (read, delete, overwrite, re-assign permissions) on every file managed within the extension.
Joomla Extension - fabrikar.com - Unauthenticated remote code execution in Fabrik < 4.6.8 - An unauthenticated attacker could execute arbitrary code by using the frontend listfilter model.
Net::CIDR::Set versions before 0.23 for Perl allow memory exhaustion and malformed set ranges via unbounded IPv6 prefix lengths. The _encode method accepts any prefix length matching `(0|[1-9][0-9]*)` and passes it to _width2bits(), which builds the mask as `'1' x ($width + 8)`, one character per bit. The _inc() method then unpacks the packed mask into a Perl array of one scalar per byte, so the prefix length alone sets the allocation size: `::/100000000` builds a 100 MB string and a 12.5 milli
Threat actors have begun to actively exploit a recently patched critical security flaw in Broadcom VMware vCenter, according to new findings from QUIRSO. The vulnerability in question is CVE-2026-59310 (CVSS score: 9.8), a directory-traversal vulnerability in the VMware vCenter server that a malicious actor with network access can exploit to execute arbitrary code. Patches for the flaw were
Insufficient input validation vulnerability in the NETGEAR R7000 models allows authenticated administrators connected to the local network to make unauthorized modification to router software and functionality.
POS System developed by FitSoft has a Missing Authentication vulnerability. Unauthenticated remote attackers can directly access and operate the system.
A command injection vulnerability in certain affected NETGEAR Nighthawk devices allows a network-adjacent attacker with the ability to intercept and modify local network traffic (attacker in the middle) to compromise the confidentiality and integrity of the affected device.
Insufficient input validation vulnerability in the listed NETGEAR models allows authenticated administrators connected to the local network to make unauthorized modification to router software and functionality.
Insufficient input validation vulnerability in the listed NETGEAR models allows authenticated administrators connected to the local network to make unauthorized modification to the device software and functionality.
A stack-based buffer overflow vulnerability affects certain NETGEAR models allowing an authenticated admin user to make unauthorized modification to router software and functionality.