Search the archive

MICROSOFT ANNOUNCES NEW MOBILE-STYLE WINDOWS SECURITY CONTROLS

Microsoft announced smartphone-style app permission prompts and a "Windows Baseline Security Mode" for Windows 11.

BREAKING DOWN CVE-2026-25049: HOW TYPESCRIPT TYPES FAILED N8N'S SECURITY

CVE-2026-25049 (CVSS 9.4) exploits a type confusion flaw in n8n's expression evaluator, bypassing a prior security patch by sending object inputs instead of strings — since TypeScript types are stripped at compile-time, the runtime sanitizer never executed. Attackers use JavaScript destructuring to access the Function constructor and achieve unauthenticated RCE via public webhook endpoints, exposing stored credentials, API keys, and internal network access. Users should upgrade to n8n versions 1.123.17 or 2.5.2 and implement runtime type validation using typeof checks or a schema validation library such as Zod.

EUROPEAN COMMISSION DISCLOSES BREACH THAT EXPOSED STAFF DATA

The European Commission is investigating a breach after finding evidence that its mobile device management (MDM) platform was hacked. It has not found any evidence that mobile devices were compromised, but confirmed that attackers have accessed some staff members' personal information, such as names and phone numbers. The European Commission did not disclose how the attackers compromised its MDM system, but the attack shows similarities with other attacks on European institutions that exploit vulnerabilities in Ivanti Endpoint Manager Mobile software.

BEYONDTRUST WARNS OF CRITICAL RCE FLAW IN REMOTE SUPPORT SOFTWARE

BeyondTrust warned customers to patch a critical vulnerability in its Remote Support and Privileged Remote Access software that could allow unauthenticated attackers to execute arbitrary code via OS command injection. The vulnerability can be exploited by unauthenticated attackers without requiring user interaction by sending specially crafted client requests. BeyondTrust has secured all cloud systems and advises all on-premises customers to patch their systems manually if they haven't enabled auto updates.

THE PHANTOM FILE SYSTEM: INSIDE THE WINDOWS PROJFS

The Windows Projected File System (ProjFS) operates as a minifilter (prjflt.sys) rather than a true filesystem, using reparse points and filter communication ports to project virtual files on demand. Offensively, medium-integrity users can launch ProjFS providers without admin privileges, prevent higher-privileged processes from deleting files, and serve different file contents per process — all potential avenues for evasion and persistence. Defensively, ProjFS offers a lightweight alternative to custom minifilters for deploying canary files with rich callback data, including triggering process ID and image path, particularly relevant as Microsoft moves toward reducing kernel-mode dependencies.

HOW TO RECOGNIZE A DEEPFAKE: ATTACK OF THE CLONES

UNCOVERING THREATS THROUGH WAF LOGS: A THREAT HUNTER'S LENS

WAF logs are often overlooked in threat hunting because every request must pass through them. Defenders can look for potential reconnaissance patterns, including using non-standard HTTP methods on public pages, CMS helper endpoint exposure, systematic enumeration across multiple sites, and infrastructure-based anomalies such as TOR exit nodes. The article also includes KQL queries for practical threat detection.

PHANTOMFS

PhantomFS is a Windows ProjFS-based tool that projects virtual files whose content varies by the requesting process, serving AES-256-CBC-encrypted payloads to allowed processes (e.g., cmd.exe) while returning decoy bytes to all others. Inspired by Huntress' recent ProjFS research, the tool runs at medium integrity without admin privileges and blocks delete/rename attempts from non-allowed processes. Detection is possible via ProjFS reparse points, the PrjFlt minifilter at filter altitude 189800, and the Microsoft-Windows-ProjFS ETW provider.

EVILNEKO

EvilNeko is a project to automate container orchestration and operationalize Browser in Browser attacks for red teams.

IDONTLIKEFILELOCKS

IDontLikeFileLocks is a tool that dumps locked browser databases.

17% OF 3RD-PARTY ADD-ONS FOR OPENCLAW USED IN CRYPTO THEFT AND MACOS MALWARE

17% of third-party "skills" for the popular open-source AI tool OpenClaw, which has over 160,000 GitHub stars, are malicious. Over half (54%) of these threats target cryptocurrency users with fake wallet trackers, DEX tools, and Solana/Phantom utilities. One user account, sakaen736jih, is associated with 199 malicious skills that steal private wallet keys and deploy AMOS Stealer malware on macOS. The malware campaign has extended into corporate settings, leading Bitdefender to develop a free AI Skills Checker tool to help users vet add-ons before installing.

MY TOP 5 RECOMMENDATIONS ON OT CYBERSECURITY STUDENT UPSCALING

A guide from Lesley Carhart, aka hacks4pancakes, on the skills prospective Operational Technology (OT) students should focus on. Students should focus more on processes and systems of systems rather than hacking individual devices. Start with one process, consider safety and process continuity, since most OT engagements involve critical systems, and get comfortable with older computers. The post also includes a list of free learning resources.

ROMANIA'S OIL PIPELINE OPERATOR CONFIRMS CYBERATTACK AS HACKERS CLAIM DATA THEFT

Romania's national oil pipeline operator, Conpet, confirmed that a cyberattack disrupted its IT infrastructure.

CHINESE CYBERSPIES BREACH SINGAPORE'S FOUR LARGEST TELCOS

Chinese threat actor UNC3886 breached all four of Singapore's major telcos (Singtel, StarHub, M1, and Simba) using a zero-day firewall exploit and rootkits for persistence, prompting Singapore to deploy "Operation Cyber Guardian."

OPENCLAW INTEGRATES VIRUSTOTAL MALWARE SCANNING AS SECURITY FIRMS FLAG ENTERPRISE RISKS

OpenClaw integrated VirusTotal malware scanning into its ClawHub skills marketplace after security researchers discovered 341 malicious skills in a campaign dubbed "ClawHavoc."

FLICKR EMAILS USERS ABOUT DATA BREACH, PINS IT ON 3RD PARTY

Flickr disclosed a data breach stemming from a compromised third-party email service provider, with attackers potentially accessing names, email addresses, usernames, IP addresses, general locations, and platform activity. The company shut down access to the affected system within hours and notified data protection authorities in both Europe and the US. No passwords or financial data were reportedly exposed, though Flickr has urged users to watch for phishing attempts and review account settings.

NEW CLICKFIX VARIANT 'CRASHFIX' DEPLOYING PYTHON REMOTE ACCESS TROJAN

Microsoft identified a new ClickFix variant dubbed "CrashFix" that uses a malicious Chrome extension impersonating uBlock Origin Lite to deliberately crash victims' browsers, then tricks users into executing clipboard-copied commands via a fake security warning. The attack chain abuses the legitimate Windows utility finger.exe (renamed to ct.exe) to retrieve obfuscated PowerShell payloads, ultimately deploying a Python-based RAT called ModeloRAT that selectively targets domain-joined enterprise systems for further compromise. ModeloRAT beacons to hardcoded C2 servers over HTTP, establishes persistence via registry Run keys and scheduled tasks, and conducts network reconnaissance using native Windows commands like nltest and net use.

PAYMENTS PLATFORM BRIDGEPAY CONFIRMS RANSOMWARE ATTACK BEHIND OUTAGE

BridgePay Networks Solutions has confirmed that an ongoing outage in its payment gateway and other services is caused by a ransomware attack. The company has stated that it is confident that payment details and credit card information have not been compromised. BridgePay has not responded to questions about which ransomware gang is behind the attack.

DEOBFUSCATION AND ANALYSIS OF RING-1.IO

Researchers reverse-engineered ring-1.io, a prominent game cheat provider, revealing a sophisticated attack chain that replaces EFI boot binaries, injects into Hyper-V via VMEXIT hooks, and uses EPT-based memory redirection to hide malicious code from kernel-level anti-cheat systems. The implant operates across three privilege boundaries — guest user mode, guest kernel mode, and VMX root — using cloned page tables, CR3 spoofing, and shadow pages filled with 0xCE bytes to evade detection. The analysis also outlines multiple detection strategies, including shadow PML4E scanning, Intel Processor Trace analysis, and RWX anomaly detection under HVCI, emphasizing that Secure Boot enforcement would prevent the entire attack chain.

HOW WE PREVENTED CURSOR, WINDSURF, & GOOGLE ANTIGRAVITY FROM RECOMMENDING MALWARE

Researchers discovered that AI IDEs forked from VSCode, including Cursor, Windsurf, and Google Antigravity, inherited hardcoded extension recommendation lists pointing to Microsoft marketplace namespaces that were unclaimed on OpenVSX, the open-source alternative these IDEs actually use. An attacker could have registered these unclaimed namespaces and uploaded malicious extensions that the IDEs would proactively recommend to millions of developers based on file types or installed software. The researchers preemptively claimed the vulnerable namespaces and coordinated with the Eclipse Foundation and vendors to remediate, noting that over 1,000 developers installed their inert placeholder extensions simply because their IDE recommended them.

GOODBYE TO STATIC CREDENTIALS: EMBRACE MODERN IDENTITY PRACTICES

Static credentials are prone to leakage, can cause significant damage, and can be difficult to rotate. To prevent this, many organizations adopted interim secrets management solutions, such as AWS Secrets Manager and HashiCorp Vault. In the long term, organizations should shift to modern, short-lived credentials, such as managed identities (e.g., AWS roles), federated identities, and application-specific methods for Kubernetes and AI agents.

See the full breakdown

HERMES

Hermes is a Linux-only Mythic C2 agent written in Python, featuring 18 built-in commands for reconnaissance, file operations, and shell execution. The agent communicates via HTTP, using Mythic's EKE + AES encryption, and can be deployed as a Python script or a PyInstaller binary. It supports core post-exploitation tasks, including process listing, network enumeration, file transfer, and directory manipulation on Linux targets.

SHAREPOINTDUMPER

SharePointDumper is a PowerShell-based extraction and auditing utility that enumerates SharePoint sites a user has access to via Microsoft Graph and downloads files via SharePoint.

CLAWDSTRIKE

Clawdstrike provides runtime security enforcement for agents. It is designed for developers building EDRs and security solutions on top of OpenClaw.

'ENCRYPT IT ALREADY' CAMPAIGN PUSHES BIG TECH TO PRIORITIZE E2E ENCRYPTION

The Electronic Frontier Foundation launched its "Encrypt It Already" campaign, urging major tech companies such as Bluesky, Google, and Ring to fulfill their promises to implement end-to-end encryption by default across their platforms. The initiative highlights that many companies either offer E2EE as an opt-in feature or have delayed rolling it out entirely, leaving users exposed, particularly as AI agents increasingly access sensitive communications with less human oversight. EFF noted that several targeted companies are "highly likely" to enable these features within the year, but stressed that default-on encryption remains critical since most users never change default settings.

A LINKEDIN JOB OFFER TRIED TO INSTALL MALWARE ON MY MACHINE

The author received a LinkedIn post about a freelance opportunity to evaluate the codebase of a real estate tech platform with a legitimate-looking platform and a generous budget. The author accepted the offer and was given access to the codebase, but missed red flags such as a missed call with their Tech Manager, suspicious git history, and suspicious indicators in their contact's LinkedIn profile. Upon reviewing the repository, the author found an npm post-install script that would download a C2 backdoor, exfiltrate sensitive files, and capture keystrokes.

SIXTEEN CLAUDE AI AGENTS WORKING TOGETHER CREATED A NEW C COMPILER

Anthropic researcher Nicholas Carlini used 16 parallel Claude Opus 4.6 agents coordinating via a shared Git repository to produce a 100,000-line Rust-based C compiler capable of compiling a bootable Linux kernel across x86, ARM, and RISC-V architectures. The two-week, $20,000 experiment achieved a 99% pass rate on the GCC torture test suite but hit a practical ceiling around 100,000 lines where new fixes frequently broke existing functionality, suggesting current limits for autonomous agentic coding. The researcher noted significant human scaffolding was required — including custom test harnesses, context-aware output filtering, and time-boxing — raising concerns about developers deploying AI-generated software they have never personally verified.

wrong. 12 short episodes from Broadcom. Start watching →

APPLE IS WORKING TO MAKE CARPLAY COMPATIBLE WITH AI CHATBOTS LIKE CHATGPT

Apple is reportedly developing support for AI chatbot apps like ChatGPT, Gemini, and Claude within CarPlay, potentially allowing drivers to interact with third-party AI assistants alongside Siri.

MALICIOUS PACKAGES FOR DYDX CRYPTOCURRENCY EXCHANGE EMPTIES USER WALLETS

Malicious npm and PyPI packages published through compromised official dYdX accounts exfiltrated wallet seed phrases and deployed a RAT backdoor, marking the third supply chain attack targeting the decentralized exchange.

NORWEGIAN INTELLIGENCE DISCLOSES COUNTRY HIT BY SALT TYPHOON CAMPAIGN

Norway's domestic security agency confirmed that Chinese state-sponsored group Salt Typhoon compromised network devices in Norwegian organizations.

CRITICAL N8N FLAWS DISCLOSED ALONG WITH PUBLIC EXPLOITS

CVE-2026-25049 enables any authenticated n8n user with workflow-editing permissions to escape the JavaScript sandbox via incomplete AST-based sanitization, resulting in unrestricted RCE and access to all stored credentials, API keys, and connected cloud accounts. The flaws bypass the prior patch for CVE-2025-68613 via a type confusion vulnerability in property access key validation, with public PoC exploits now available and GreyNoise reporting 33,000+ probing requests against exposed n8n endpoints. Organizations should immediately update to n8n versions 2.5.2 or 1.123.17, rotate the N8N_ENCRYPTION_KEY and all stored credentials, and restrict workflow editing to trusted users only.

CVE-2025-22225 IN VMWARE ESXI NOW USED IN ACTIVE RANSOMWARE ATTACKS

CISA confirmed that ransomware gangs are actively exploiting CVE-2025-22225 (CVSS 8.2), an arbitrary kernel write vulnerability in VMware ESXi that enables sandbox escape from the VMX process, now added to the KEV catalog as used in ransomware campaigns. Huntress attributed a related exploit chain to Chinese-speaking threat actors using the MAESTRO orchestrator toolkit, which chains CVE-2025-22224, CVE-2025-22225, and CVE-2025-22226 to achieve full VM escape and deploy the VSOCKpuppet backdoor for persistent hypervisor control. Organizations running unpatched ESXi should apply VMSA-2025-0004 immediately and audit for signs of unauthorized kernel driver loading or VSOCK-based lateral movement from guest VMs.

DATA BREACH AT GOVTECH GIANT CONDUENT BALLOONS, AFFECTING MILLIONS MORE AMERICANS

A ransomware attack on Conduent in January 2025 exposed highly sensitive data, including names, Social Security numbers, medical details, and insurance information, for tens of millions of Americans across multiple states. Conduent, which supports government healthcare programs for over 100 million people, is still tallying victims and slowly issuing breach notifications. It has offered few specifics about the incident's full scale.

AUTHENTICATION DOWNGRADE ATTACKS: DEEP DIVE INTO MFA BYPASS

IOActive demonstrated an authentication downgrade attack using Cloudflare Workers as a transparent reverse proxy to intercept Microsoft Entra ID login flows, modifying the JSON configuration and injecting CSS to hide FIDO2/WebAuthn options and force victims to fall back to phishable MFA methods, such as push notifications. The technique leverages trusted CDN infrastructure with valid TLS certificates and zero forensic footprint, enabling session token capture (ESTSAUTH cookies) while authentication logs appear completely legitimate. Organizations with "mixed mode" MFA policies should enforce FIDO2-only Conditional Access rules without fallback methods and alert when users unexpectedly switch from hardware keys to weaker authentication factors.

HOW LLMS FEED YOUR RE HABIT: FOLLOWING THE USE-AFTER-FREE TRAIL IN CLFS

AI tools helped a security researcher track down a serious memory bug in a Windows file system driver. By comparing an old and a new version of the driver, the author identified a risky change and then used timing tricks to trigger a bug in which freed memory is reused. Connected to Ghidra, LLMs explained confusing functions, named important structures, and highlighted where the bug lived, so the researcher could focus on logic instead of wading through code.

HANDS ON WITH AWS BOTTLEROCKET: EVALUATING THE SECURITY OF AMAZON'S HARDENED OS

AWS Bottlerocket is a hardened OS designed for running Kubernetes nodes that includes several layers of protection, such as a read-only root filesystem, an always-enabled, restrictive SELinux profile, no package managers, language interpreters, or command shells, as well as removing many other common privilege escalation tools. This article compares an Ubuntu node and a Bottlerocket node with three common container escape techniques and found that Bottlerocket prevented all three of them, while they all succeeded on the Ubuntu node.

patch. Read the blog

K8S-CUSTOM-DETECTIONS

k8s-custom-detections provides a curated collection of Falco detection rules, audit policies, and reproducible attack scripts for identifying real-world Kubernetes attack techniques, including RBAC abuse, service account token theft, CoreDNS manipulation, etcd unauthorized access, and certificate forgery. The repository includes ready-to-deploy lab setup scripts for Ubuntu 24.04 with Helm-based Falco integration and individual test scenarios that simulate attacker behavior against each detection category. Useful for security teams building Kubernetes runtime detection capabilities or validating existing monitoring coverage against common cluster compromise techniques.

ZEN AI PENTEST

AI-Powered Penetration Testing Framework with automated vulnerability scanning, multi-agent system, and compliance reporting.

NULLIFY (PRODUCT LAUNCH)

Nullify offers an AI workforce for product security that autonomously detects, triages, validates, prioritizes, and remediates vulnerabilities in codebases and cloud environments. It integrates with tools like Jira, generates exploit proofs, and automates issue fixes via merge-ready PRs, replacing manual security engineering.

THREAT ACTORS HIJACK WEB TRAFFIC AFTER EXPLOITING REACT2SHELL VULNERABILITY

Datadog Security Labs observed threat actors exploiting the React2Shell vulnerability (CVE-2025-55182) in React 19 server components to pivot into NGINX web servers managed via Boato Panel, deploying automated toolkits that modify configuration files to hijack web traffic. Targets predominantly include Asian organizations across .in, .id, .gov, and .edu domains, with attackers using compromised servers for traffic fingerprinting, malware delivery, and credential phishing via redirect pages. Defenders should monitor NGINX configuration file integrity, ensure React server components are patched, and audit web server access controls to detect unauthorized configuration changes.

1.5 MILLION AI AGENTS ARE AT RISK OF GOING ROGUE

SENATOR PRESSES AI TOY COMPANY ON CHILD DATA PRIVACY

Senator Maggie Hassan has shared a letter with bondu, a company that makes AI conversational toys for children, about their data privacy practices. This inquiry comes in the wake of a finding by security researchers that bondu accidentally exposed chat transcripts and personal data through a publicly accessible portal. Hassan's letter specifically asks bondu who has access to the portal, what steps have been taken to prevent future incidents, and whether AI coding practices were used to develop the portal.

CYBERSPY GROUP HACKED GOVERNMENTS AND CRITICAL INFRASTRUCTURE IN 37 COUNTRIES

A nation-state group tracked as TGR-STA-1030 has breached at least 70 organizations in 37 countries, targeting parliaments, telecoms, law enforcement, and critical infrastructure for espionage.