Sophos Urges Rules at Sensitive Uploads Before AI Judges Intent
Endpoint AI can flag unusual access, but a familiar-looking transfer may still be harmful. Put explicit policy at the upload and measure the review queue.
Read article →2026
Nulltap reporting published in 2026.
Google BoringSSL Fixes Two Certificate Name-Constraint Bypasses
Two legacy BoringSSL name checks were fixed in March; a URI case remains under coordination. Find the validator your product actually uses.
Read article →
UAT-6275 uses a WinRAR flaw to plant a shell in CIS government attacks
Cisco Talos traces four delivery routes across nine countries. Hunt Startup-folder changes and unexpected PowerShell, then update affected Windows archive tools.
Read article →
FBI Disrupts Laptop Farms Behind North Korean Remote Hires
CrowdStrike says fake employees used U.S. homes to run company laptops remotely. Inspect shipping changes, remote-access tools and hiring records together.
Read article →
Fortinet FortiSIEM Flaw Lets Remote Attackers Run Code
Crafted packets can cross a file-path restriction in FortiSIEM 7.5.0. Upgrade to 7.5.1 and limit access to TCP/27918 while scheduling the change.
Read article →
Bank Intruder Targets Update Server After Third-Party Gateway Entry
At a European bank, compromised credentials opened a third-party gateway route; months of scanning and password guessing preceded a blocked attempt against WSUS.
Read article →
Attackers Hid PEERSTITCH in Fortinet Firewalls After Manager Breach
A stolen FortiManager archive helped attackers reach managed FortiGates; a memory-only implant linked logging and VPN processes. Check revisions and rotate trust material.
Read article →
Microsoft and NVIDIA Map Open-Weight Model Risk to Enterprise Deployments
The joint guidance separates model evaluations from the permissions, data access and tools of the application that uses a model.
Read article →
Volexity Volcano Recovers Browser Exploit Traces After Cleanup
A replayed Chrome-to-Windows intrusion left encoded script in free memory. Capture RAM early, then use host artifacts to guide wider collection.
Read article →
Fortinet CLI Flaw Lets Privileged Attackers Change Appliance Files
CISA added a flaw affecting four Fortinet products to KEV. Match installed releases to Fortinet's fixes, then examine privileged CLI access and file changes.
Read article →
ConnectWise Backup Credential Opens Graph Access in 162 Tenants
An abused vendor-held Cloud Backup credential obtained app-only Graph tokens. Find the service principal, disable sign-in, then inspect roles and added secrets.
Read article →
Cisco Email Gateway Attachment Flaw Gives Attackers Root Access
Exploited file-write flaw reaches gateways using File Reputation and older scanner tools. Check each appliance's logs and scanner version before choosing an update or rebuild.
Read article →
CL-UNK-1211 Plants PHP Web Shells on Government Linux Servers
Attackers changed PHP configuration to reload a web shell, stole passwords through a sudo alias and searched government mailboxes for social media access.
Read article →
Attackers Briefly Hijacked AOMEI Updates to Plant Backdoors
UNC7017 changed two AOMEI configuration files for minutes at a time. Hunt updater-launched files and scheduled tasks even after the vendor restored its servers.
Read article →
ARMORY CASTLE Uses BeyondTrust Zero-Day to Keep Root Access
The actor exploited Remote Support, installed a root backdoor and reached managed endpoints. Patch BT26-05, then investigate appliance persistence and sessions.
Read article →
Fortinet FortiSandbox Flaw Lets Remote Callers Bypass GUI Login
Crafted HTTP requests can bypass FortiSandbox management login. Check appliance and PaaS versions, upgrade affected branches and restrict GUI reachability.
Read article →
TLS Libraries Disagree on Which Signed Certificates Pass Name Restrictions
The same signed certificate can pass one validator and fail another. Test the actual TLS path, update affected libraries and keep issued names explicit.
Read article →
Komodo Log Access Lets Restricted Users Run Commands on Managed Hosts
Komodo Periphery passes some restricted-user requests to a shell on managed servers. Upgrade agents to 2.3.0 and review Logs and Execute grants.
Read article →
Lowcoder Flaw Lets One Tenant Run Another Tenant's Saved Queries
An admin or developer in one organization can reach another's Query Library on a shared instance. Restrict the affected API while a security update is pending.
Read article →
Bisheng SSO Issues Admin Tokens Without Checking the Caller
The legacy SSO route issues a JWT for a supplied username and remains unpatched; separate cross-user assistant and workflow flaws have release-specific fixes.
Read article →
Cisco Email Hardening Fixes Five Flaw Groups, Including Two Exploited Issues
Cisco updates Secure Email Gateway and Secure Email and Web Manager across five weakness groups. Compare each appliance's build with its product-specific fixed release.
Read article →
Palo Alto Networks PAN-OS Flaw Lets Attackers Run Code as Root
Limited exploitation targets authenticated access to PAN-OS management. Compare installed hotfixes, restrict the admin path and inspect three vendor-listed IPs.
Read article →
YAM Store sells Microsoft and Google cloud access to phishing crews
Microsoft traces a broker reselling cloud tenant access and storage to phishing kits. Examine app credentials, privileged grants and hosted files in affected tenants.
Read article →
Microsoft Patches Exploited Windows File Warning Bypass
The October updates address an exploited Windows Shell bypass and separate unauthenticated DHCP Server code execution. Check installed builds on endpoints and DHCP hosts.
Read article →
Fake WatchGuard VPN download takes passwords before serving the real client
A lookalike download page collects the VPN server and credentials twice, then starts a signed WatchGuard installer. Check browser history and VPN logins.
Read article →
Microsoft Windows Shell flaw lands in CISA exploited catalog
CISA flags active exploitation of a Windows Shell security bypass. Compare affected desktop and server builds with Microsoft's CVE record and prioritize updates.
Read article →
Shopware media uploads expose admin tokens to stored XSS
An uploaded HTML file can run in a Shopware administrator's browser. Restrict media creation or block HTML uploads while checking who can reach existing media.
Read article →
Ivanti VPN Zero-Days Let UTA0576 Reach Root and Steal Credentials
A patched-at-the-time Connect Secure VPN fell to a three-flaw chain. Upgrade affected branches and investigate memory, LDAP traffic and movement from the appliance.
Read article →
GitLab fixes CI regex flaw that permits server-side code execution
A crafted CI configuration and UTF-8 repository path can reach a vulnerable regex compiler on self-managed GitLab. Upgrade the affected 19.2, 19.3 and 19.4 branches.
Read article →
Microsoft Tests Post-Quantum Certificates Before Enterprise Rollout
A closed TLS pilot exposes the certificate-chain work ahead. Map issuers and validators, then test ML-DSA certificates across real application paths outside production.
Read article →
Fake event invites target Taiwan researchers’ Google accounts
Talos traced event invitations and altered poster QR codes to a live Google sign-in relay. Inspect the actual registration destination and review affected accounts.
Read article →
Apex Accelerator's fake audits deliver malware through npm tools
Two Apex npm tools fetch executables and attacker-selected commands; one also sends coding-agent conversations to an Apex server. Find the tools and revoke exposed secrets.
Read article →
Thelia back-office flaws expose store servers to code execution
CERT/CC details code-execution paths in Thelia 2.6; a separate Thelia 3 email-template flaw has a module fix. Restrict admin access and check permissions.
Read article →
Cisco Talos Measures Malware Prompts That Mislead AI Scanners
Plaintext instructions in malware samples shifted some model triage toward benign. Search extracted strings and keep sample text out of the scanner's instruction channel.
Read article →
Huntress kills a ClickFix Run chain before retrieved text executes
In a Windows ClickFix case, endpoint-side termination stopped repeated command pastes; check enforce mode and the process tree before treating an alert as a block.
Read article →
CISA lists exploited Strapi filter flaw that exposes private fields
Strapi user filters can disclose reset tokens. Find vulnerable releases, inspect GET requests for private-field probes and move to a patched branch.
Read article →
CISA flags exploited ProFTPD flaw that exposes server files
ProFTPD's mod_copy flaw lets remote attackers read and write arbitrary files. Check reachable FTP services, command logs and distribution patches.
Read article →
Cisco patches Nexus MPLS flaw that permits root-level code execution
An enabled MPLS OAM feature exposes affected Nexus switches to crafted echo-requests; check feature state, then disable it or move to fixed NX-OS.
Read article →
Cisco NGOAM flaws expose Nexus switches to root code execution
Enabled NGOAM exposes a packet-processing flaw on standalone Nexus 3000 and 9000 switches; two more flaws require VXLAN overlay or SRv6 features.
Read article →
AhsayCBS Intruders Install Webshells and Hide a Crypto Miner
A backup management server became a SYSTEM-level foothold. Restrict its web interface and hunt for a fake Edge service, JSP shells and miner traffic.
Read article →
Cisco NX-API flaw exposes switch management to root execution
Enabled NX-API on affected Nexus switches exposes an unauthenticated request path; UCS 6300 fabric interconnects have a separate credential-dependent path.
Read article →
CISA Flags Exploited BIND Flaw That Knocks DNS Servers Offline
A crafted TKEY query can stop vulnerable BIND DNS servers before access controls run. Check the running build and install the vendor's patched package.
Read article →
PaperCut Emergency Patch Left a Second Path to Server Code Execution
watchTowr traced a new authentication bypass and Scan-to-Fax code-execution chain through an interim PaperCut build. Check the running version and past access to its web interface.
Read article →
Citrix NetScaler Intruders Poison Logs to Install a Web Shell
Unit 42 traced an exploited NetScaler command-injection chain through two appliance logs into a disguised PHP shell. Preserve logs and inspect the web-server configuration.
Read article →
Cisco Fixes Six NX-OS Flaw Groups Across Switches and Fabric Interconnects
Cisco's NX-OS hardening release covers MDS, Nexus and UCS fabric interconnects. Match device family and management mode to the first fixed build.
Read article →
Cisco Fixes Four License On-Prem Flaw Groups in 10-202609
Cisco groups multiple License On-Prem flaws under four CVEs. Inventory the former SSM On-Prem too, migrate older branches, and verify the fixed release.
Read article →
CISA Flags Exploited Apache Struts Flaw That Executes Server Code
A remote execution path depends on Dynamic Method Invocation. Identify reachable Struts 2 apps, disable that feature or move to an Apache-listed fix, and triage exposure.
Read article →
Microsoft FORGE Finds 140 Windows CVEs, Submits 155 Open-Source Reports
Microsoft's research shows how bug volume can overwhelm validation and release work. Check affected dependency builds and carry reproducible findings through a shipped fix.
Read article →
CISA flags exploited ONLYOFFICE Docs flaw that exposes servers to code execution
CISA lists an exploited image-upload traversal in ONLYOFFICE Document Server before 5.6.3. Find older JWT-enabled servers, update them and check their exposure.
Read article →
Akira Ransomware Leaves a File-Share Trail After Late EDR Deployment
Huntress reconstructed an Akira attack from Shellbags, ransomware logs and shadow-copy events after its endpoint agent arrived too late to record the launch.
Read article →
Palo Alto Networks traces blockchain C2 through poisoned dependencies
Unit 42 maps contract reads, transaction data and zero-value transfers used by poisoned packages. Correlate runner processes with blockchain traffic and restrict unneeded egress.
Read article →
Atlassian file-read flaw exposes web-root secrets in eight products
An unauthenticated file read reaches known paths inside eight Atlassian server products. Match installed branches to fixes and review decoded access logs.
Read article →
Microsoft Power BI phishing plants two rogue ScreenConnect clients
A Power BI report link sent targets to a fingerprinting page and a delayed installer. Hunt for both remote clients and the recurring repair task.
Read article →
Foxit PDF form flaws expose Reader users to code execution
Two crafted-PDF flaws affect Foxit Reader and Editor. Check both CVEs against installed platforms and releases, then verify Foxit updates on endpoints.
Read article →
Microsoft urges 24-hour fixes for critical on-premises systems
Microsoft urges faster fixes for critical on-premises systems. Check patch deployment logs, installed-update status and Baseline Security Mode configuration exceptions.
Read article →
Iranian Actors Trigger Malware When Visual Studio Opens a Coding Test
A fake Dubai Airports coding test ran from a Visual Studio project before compilation. Inspect the project target and developer-host staging path, then isolate matches.
Read article →
NeedyMantis Uses Sideloaded DLLs to Maintain Windows Access
Microsoft found a selective post-compromise implant paired with legitimate software and a same-name archive. Hunt the loader, archive and C2 together, then contain affected hosts.
Read article →
Attackers use MSP360 and ScreenConnect to keep access after phishing
A disguised signed MSP360 installer gave intruders one management channel, then installed ScreenConnect for another. Hunt the service-to-installer chain and restrict unapproved agents.
Read article →
Cisco Talos maps RMM, decoys and agent controls to disrupt attackers
A legitimate remote-support agent can give a ransomware operator persistent access. Inventory approved tools, watch credential-access behavior and make agent sessions stoppable.
Read article →
Citrix NetScaler flaw threatens service as CISA lists exploitation
CISA lists CVE-2026-88779 as exploited. Record each ADC and Gateway's installed version and internet exposure; upgrade affected builds and verify the result.
Read article →
Arista VeloCloud Orchestrator flaw exposes privileged host functions
CISA lists the on-prem flaw as exploited. Check Edge certificate authentication, web-interface reachability and the affected release train before upgrading.
Read article →
Signed UEFI shells bypass Secure Boot before the OS starts
CERT/CC identifies trusted diagnostic shells with memory-write commands. Match the binary and firmware trust lists, then install vendor updates and verify DBX revocation.
Read article →
Zyxel GS1900 switches face exploited CGI command-execution flaw
CISA lists a LAN-reachable GS1900 CGI overflow as exploited. Match switch model and firmware to Zyxel’s fixes, then restrict management access.
Read article →
Readwise Reader Android XSS Persists in Metadata After SVG Patch
Readwise Reader 8.10.1 fixes an SVG sanitizer flaw, while two metadata XSS records still include that build. Check Android versions and limit untrusted imports.
Read article →
Star Blizzard Sends RedFlick Archives to Plant Windows Tasks
Reply-stage archives lead to RedFlick scheduled tasks and CosmicPulse. Hunt the task names, trace the archive, and verify senders through a known channel.
Read article →
WSO2 API Flaw Draws Conflicting Accounts as CISA Confirms Exploitation
CISA describes file upload and possible code execution; WSO2 describes a JWT bypass and account takeover. Check product builds and reconcile the attack path with WSO2.
Read article →
CISA Flags Exploited Zammad Session Flaw Leading to Code Execution
CISA lists an exploited session flaw with code execution as the Zammad user. Zammad says 6.5 and older are exploitable; move those installations to 7.2.0 and triage exposure.
Read article →
CISA Lists Exploited F5 BIG-IP APM Flaw Allowing Code Execution
The exposed configuration is an APM OAuth Authorization Server virtual server. Request F5’s interim iRule for triage, then install the applicable final fix.
Read article →
Cisco Talos Uses CAIRN to Hunt AI Malware From Sample Metadata
CAIRN groups malware candidates by AI-related metadata, then separates artifact presence from behavior and confirmed families. Triage noisy matches before calling an incident.
Read article →
Storm-2570 Reuses Remote Access to Steal Data and Deploy Ransomware
Microsoft tracked the same affiliate through four ransomware ecosystems. Hunt renamed MeshAgent services, domain-controller copies and cloud-transfer tooling before encryption.
Read article →
Norwegian Cruise Line Ship Doors Accept Cloned Keycards
Some shipboard access readers accept a copied card identifier. Identify UID-only doors, shield issued cards and add supervised access to sensitive areas while seeking a fix.
Read article →
Microsoft Brings Sensitive AI Upload Blocking to Entra Networks
Purview classifies files and text while Entra enforces network policy for employee and on-behalf-of agent traffic. Confirm forwarding and an active blocking rule before relying on it.
Read article →
CISA Flags Exploited Zammad Flaw That Elevates Service Account to Root
A local zammad account can escalate to root. CISA lists exploitation; the vendor disputes the reported scope and recommends upgrading to 7.2.0.
Read article →
IBM Turbonomic Operator Grants Cluster-Wide Access to Secrets
Prometurbo's service account could read secrets outside its own namespace. Reinstall the fixed agent and test the operator's effective permissions.
Read article →
MikroTik RouterOS Rekey Error Exposes Routers With Open SSH
CISA lists a RouterOS SSH rekey flaw as exploited. It lets unauthenticated clients reach a command path and can chain with a second flaw; update and restrict SSH.
Read article →
HP BIOS Flaw Lets Kernel Attackers Write to Protected Firmware Memory
An HP-specific InsydeH2O SMM handler accepts unsafe memory parameters from a local kernel attacker. Check HP PCs for exposure and install a model-matched BIOS update when HP lists one.
Read article →
Fortinet FortiMail Attackers Write Files Through Its Web Interface
CISA lists an exploited FortiMail arbitrary-file-write flaw. Fortinet urges operators to disable IBE or restrict management access while branch upgrades are forthcoming.
Read article →
ViewSonic ViewBoard Flaws Expose Screens and Open a Code Path
Three unauthenticated vCast endpoints expose screen images, APK delivery and service input on shared networks. Isolate ViewBoards and check who can reach them.
Read article →
Imprivata EAM Keeps Appliance Keys Trusted Until Redeployment
EAM through 26.2.6 has no supported RSA key-pair rotation. Protect key copies in backups and snapshots, and plan redeployment if the private key is exposed.
Read article →
Storm-3068 Steals Kubernetes Credentials via Microsoft Azure DevOps
After a password reset, Storm-3068 used an authorized Azure DevOps pipeline to collect Kubernetes credentials. Trace pipeline changes and restrict resource access.
Read article →
Zimbra Mail Clusters Face Secret Theft After Server Intrusions
Microsoft traces a mail-server intrusion from an optional SNMP notification path to root persistence, shared authentication secrets and peer mailbox nodes.
Read article →
Cisco SD-WAN Manager Flaw Gives Attackers Admin Access
An exploited URI-encoding bypass reaches the management API as admin. Check Cisco's service-proxy and server logs, restrict exposure, and install a fixed release.
Read article →
Antino Backdoor Uses Microsoft 365 for Espionage Tasking
Talos traces a phishing-to-Windows backdoor chain affecting Asian institutions. Hunt the script host and sideloaded DLL, then restrict script execution.
Read article →
Kotaemon Lets Signed-In Users Read and Delete Other Users' Chats
An ownership check is missing from four multi-user chat handlers through v0.12.0. Limit account access and review cross-user conversation activity while awaiting a verified fix.
Read article →
GTIG Counts 141 Exploited Flaws as CVE Disclosures Double
GTIG recorded 141 disclosed and exploited vulnerabilities through August 2026. Prioritize reachable edge systems and verify the access conditions on affected AI gateways.
Read article →
CISA Flags Exploited SharePoint Code Injection for Authorized Users
Three on-premises SharePoint editions have affected build ranges. Match each server to its edition's update threshold and triage suspected compromise.
Read article →
NetScaler Intruders Hide Web Shells Behind Icons and Tunnel Into Networks
Mandiant traces exploited NetScaler packet-engine crashes to disguised PHP shells and an internal proxy. Hunt the config edits and appliance egress, then patch both flaws.
Read article →
ShinyHunters Bypass PeopleSoft WAF Rules and Plant Web Shells
An encoded request reaches PeopleSoft's vulnerable EMHub servlet past literal path filters. Patch PeopleTools and search every WebLogic node for encoded paths and new JSPs.
Read article →
Authlib accepts unsigned JWS payloads as verified
An empty signatures array skips Authlib's verification loop. Locate callers of the JSON JWS deserializer and reject unsigned input before trusting its payload.
Read article →
Check Point Management Servers Face Script-Upload Attacks
Check Point reports customer attacks through an unauthenticated management-server flaw. Match Jumbo Hotfix takes, restrict TCP/19009 and inspect CPM logs.
Read article →
AWS Quarantines an Exposed IAM Key While CloudTrail Names the User
A public-key exposure test reached AWS quarantine in 10 seconds. Alert on the IAM policy attachment, then investigate the key and follow the Support case.
Read article →
Mandiant Details How Shared CI Caches Expose Release Builds
A pull request can plant dependencies in a shared build cache for a privileged release job. Separate cache writes by branch privilege and constrain publishing identities.
Read article →
Attackers Exploit Two NetScaler Flaws That Permit Code Execution
CISA lists two exploited NetScaler flaws. Match ADC and Gateway builds to their release branches, then preserve evidence on appliances suspected of compromise.
Read article →
Storm-3168 Deletes Azure Storage Accounts With Stolen Service Principals
Two compromised service principals drove a rapid Azure deletion campaign. Trace their ARM operations, rotate exposed secrets and verify recovery-resource locks.
Read article →
Sogou's Link Handler Let UNC3569 Install GRAYRABBIT
A crafted Sogou URI reached an old, unsandboxed Chromium renderer. Verify the fixed build, then hunt the process, file, and network chain.
Read article →
BlueMoon Exploit Kit Breaks Out of Chrome to Deploy Espionage Malware
Four espionage groups used a shared Chrome-to-Windows exploit chain. Build checks show current exposure; campaign artifacts support a separate compromise hunt.
Read article →
RubyDoc Builds Let Malicious Gems Run Code on Shared Servers
More than 100 malicious gems used RubyDoc documentation builds to execute supplied code. Maintainers should review package and account changes tied to legacy API keys.
Read article →
Attackers Exploit ScreenConnect Flaw to Run Files Without Host Approval
CISA confirms exploitation of CVE-2026-84869. ScreenConnect clients before 26.6.5 can transfer and run files through active sessions without authorization or host confirmation.
Read article →
Attackers Exploit GitLab Flaw to Read Server Files Without a Login
An exploited GitLab flaw exposes arbitrary server files through the commits API. Self-managed operators should install 19.1.8, 19.2.6, or 19.3.2 immediately.
Read article →
BigBear Phishing Bypasses Microsoft 365 MFA at 258 Organizations
CloudSEK found BigBear 2.0 stealing Microsoft 365 session cookies after MFA. Revoke sessions, rotate credentials and require phishing-resistant sign-in.
Read article →
AOMEI Backup Driver Lets Local Users Rewrite Physical Disks
AOMEI Backupper 8.4.0 exposes physical-disk writes to local users. Without Secure Boot, attackers can plant UEFI code that runs before Windows defenses.
Read article →
Check Point VPN Flaws Let Remote Attackers Run Code on Gateways and Servers
Two certificate-processing flaws reach gateways and management servers. Fixed takes and LivePatch output give operators a direct protection check.
Read article →
Cisco Firewall Manager Breaches Led to Credential Theft and Ransomware
Attackers used Cisco Secure FMC flaws to reach root, steal credentials, tunnel into networks, deploy Cyclops Blink, and encrypt selected endpoints with Qilin.
Read article →
Attackers Exploit Kestra Login Bypass That Reaches Root
Kestra OSS versions through 1.3.20 let unauthenticated requests create workflows and run commands as root inside workers. CISA confirms exploitation.
Read article →
Google Patches Exploited Chrome Flaw Across Desktop and Android
Chrome 153 fixes exploited V8 out-of-bounds write CVE-2026-87491 across desktop and Android. Fleet owners should verify running builds; Google published no attack indicators.
Read article →
Attackers Exploit Fortinet Flaw That Lets Crafted Packets Run Code
CISA confirmed exploitation of CVE-2025-25249 across FortiOS, FortiSwitchManager and FortiSASE, with federal remediation due September 12.
Read article →
CISA Links WatchGuard Firebox Code-Execution Flaw to Ransomware Campaigns
CISA now links CVE-2025-14733 to ransomware. Patch exposed Fireboxes, check WatchGuard's indicators, and rotate locally stored secrets after confirmed activity.
Read article →
Attackers Exploit NetScaler Flaw to Bypass Authentication
Customer-managed VPN and AAA appliances need branch-specific fixed builds; configuration determines exposure on later NetScaler releases.
Read article →
Sliver Toolkit Scripts Credential Theft and Persistent Domain Access
A recovered Sliver kit shows how one operator scripted Domain Admin access, disabled defenses, stole credentials, and hid rotating command servers behind Ethereum.
Read article →
Two Windows Zero-Days Give Local Attackers SYSTEM Privileges
The flaws require a local foothold, affect different Windows release families, and leave defenders with fixed-build checks but no CVE-specific compromise indicators.
Read article →
Coder Registry Hack Sent Credential-Stealing Code to Developers
Attackers made Coder's trusted registry serve malicious Terraform modules. Operators have a 14-hour exposure window, concrete indicators, and urgent credential work.
Read article →
Cloud Intruders Automate Mass Credential Harvesting in Under Six Hours
Google observed attackers move from cloud compromise to agent-enabled mass credential harvesting in under six hours, shrinking the time defenders have to respond.
Read article →
JSCeal Replays Browser Sessions to Steal Fresh Google Tokens
Recovered JSCeal code replays stolen cookies and passwords through a headless browser to obtain fresh Google OAuth tokens and exposes concrete Windows hunt artifacts.
Read article →
Public Telerik Exploit Lets Attackers Run Code on ASP.NET Servers
A public Telerik UI exploit chains two cryptographic oracles to unsafe type loading. Upgrade ASP.NET AJAX to 2026.2.708 and check IIS for post-exploit activity.
Read article →
N-able Patches N-central Flaw That Lets Attackers Run Code Before Login
Every on-premises build before 2026.3.1.14 needs Hotfix 4. N-able's records conflict on exploitation, so MSPs should patch and review RMM activity.
Read article →
Compromised npm Code Generator Steals Developer and Cloud Credentials
Ten malicious OpenAPI React Query Codegen releases ran on install and carried credential-stealing code with valid provenance. Teams must scope lockfiles, rebuild hosts and rotate credentials.
Read article →
Claude Ports WAGO PLC Exploit and Bricks Test Controller
Forescout ported a pre-auth PLC exploit with Claude, then bricked the device during an implant attempt. Restrict FTP and monitor crashes and outbound traffic.
Read article →
ZBT Router Firmware Gives Remote Attackers Root Access
Two factory firmware implants expose white-label ZBT routers through an open WAN service and an unauthenticated phone-home channel.
Read article →
LiteLLM Flaw Lets Read-Only Accounts Steal AI Gateway Secrets
Zenity saw file-read probes matching CVE-2026-35029. LiteLLM advises upgrading to 1.83.0 or later; defenders should hunt configuration changes and rotate exposed secrets.
Read article →
MikroTik Patches RouterOS and Adds Compromise Check
CERT Polska confirms attacks through internet-exposed SSH. Install a fixed build, restrict management access, and investigate the published log and account indicators.
Read article →
Magento Zero-Day Lets Attackers Backdoor Fully Patched Online Stores
StyleSmuggler turns poisoned Magento logs into server-side code execution, then hides a persistent implant outside the shop's webroot. Adobe has not issued a fix.
Read article →
Attackers Breach JetBrains Cloud Coding Service and Expose Credentials
JetBrains confirmed that an unpatched TeamCity flaw exposed Cadence users’ code and secrets. Former users should rotate credentials and review connected systems.
Read article →
AI Agents Help Ransom Attacker Seize Enterprise Root Credentials
Unit 42 traced an AI-assisted ransom intrusion across web, repository, secrets, CI/CD and cloud systems in less than 10 hours. These behaviors can reveal the loop.
Read article →
D-Link Router Flaws Let Local Users Seize Admin Access
D-Link fixed a DIR-X1860Z flaw that lets a local-network user set a new admin password; the similar DIR-X1860 has no update path.
Read article →
Attackers Stole Philippine Nuclear Records Through an Old ownCloud Flaw
Recovered tooling shows how forged WebDAV requests exposed nuclear records and credentials, with specific version, request-pattern, and signing-key checks for defenders.
Read article →
Chinese-Speaking Operator Uses AI Agents in Government and Education Intrusions
Five exposed SecFlow workspaces linked Claude, Qwen and DeepSeek to Asian government and education intrusions involving credential theft, webshells, and implants.
Read article →
Fake Microsoft Teams Support Calls Plant Persistent Node.js Backdoors
Microsoft observed fake IT support sessions progress from Teams remote control to a persistent JavaScript implant and WinRM movement toward domain controllers.
Read article →
Google Patches Exploited Chrome V8 Flaw Across Desktop Fleets
Chrome 152.0.7977.82/.83 fixes an exploited V8 type-confusion flaw. Google has not disclosed the attackers, targets, exploit chain, or scale.
Read article →
Ransomware Server Exposes Credential Theft and EDR-Killer Toolkit
A server tied to The Gentlemen exposed TukTuk C2, a credential-stealing prompt, EDR-killer research, exfiltrated Jira data, and healthcare credentials.
Read article →
Leaked AWS Administrator Key Funds Unauthorized AI Model Access
Fortinet traced one stolen AWS administrator key through a new IAM user, Marketplace agreements, and billable Bedrock model calls.
Read article →
Attackers Used METR’s Exposed API Key for Credits Worth $600,000
A fail-open agent dashboard exposed one public-model key for three weeks, while noisy usage signals and absent spend limits delayed recognition.
Read article →
Malicious Chrome and Edge Extensions Steal Wallet Secrets and Passwords
Socket found 19 browser extensions using automatic updates, rotating command servers and injected modules to steal wallet secrets, sessions and passwords.
Read article →
Attackers Exploit Two SonicWall VPN Flaws; One Allows OS Commands
SonicWall confirmed active attacks against two SMA1000 flaws and told customers to hotfix, seek an IoC review, and rebuild systems when compromise is found.
Read article →
BREEZE COMET Hijacks Brazilian Payment Systems for Fraud
Mandiant traced BREEZE COMET from vishing, rogue branch hardware and stolen cloud credentials to payment APIs used for hundreds of fraudulent transfers.
Read article →
OpenAPI Code Generator Worm Hijacks Trusted Publishing
Ten malicious npm releases used trusted publishing, two install-time launch paths, and a credential-stealing worm. Defenders must isolate hosts before rotating tokens.
Read article →
Attackers Exploit JFrog Artifactory to Create Administrator Tokens
CISA confirms exploitation of a critical Artifactory flaw and requires federal forensic triage. Self-hosted operators have six fixed-version floors to verify.
Read article →
Malicious Repositories Make Amazon Kiro Send Local Secrets Off-Device
A crafted workspace can steer Kiro 0.7.45 from reading a local secret to placing it in a Powers registry request. Amazon fixed the reported behavior in 0.8.140.
Read article →
Fire Ant Hijacks Cisco Routers to Steal Credentials and Erase Evidence
Fire Ant hid tunnels on Cisco routers, injected TACACS servers, and planted Linux backdoors. Sygnia's artifacts support checks of routers, authentication servers, and Linux hosts.
Read article →
CISA Orders Forensic Checks for Exploited Microsoft SQL Server Flaw
CISA confirmed exploitation of a 2019 SQL Server flaw and now requires affected systems to be patched after evidence preservation and forensic triage.
Read article →
Attackers Exploit PaperCut Servers to Install Remote Access
PaperCut confirms active exploitation of NG and MF servers. Release 3 patches the two-flaw chain; new log and service indicators help defenders investigate exposed hosts.
Read article →
Hackers Use Nearly 2,000 WordPress Sites for Theft and Ransomware
StopAndProtect turns hacked WordPress sites into malware hosts, command servers, and stores for stolen files before selective ransomware deployment.
Read article →
Dysphoria Compromised 296,000 IoT Devices for DDoS and Proxy Abuse
Shadowserver's retrospective Dysphoria report identifies about 296,000 compromised IoT devices and gives network owners evidence to locate and rebuild affected systems.
Read article →
WordPress Extension Flaws Expose Admin Accounts and Web Servers
Five newly fixed WordPress flaws expose conditional paths to admin takeover or server code execution; defenders should verify six component versions and review AJAX and account activity.
Read article →
Isolated-vm Flaw Lets Guest Code Hijack the Host Process
A double-read type confusion lets guest JavaScript corrupt host memory; upgrade isolated-vm 6.x to 6.2.0 or 7.x to 7.0.1.
Read article →
CISA Flags Exploited Linux IPv6 Flaw That Lets Local Users Gain Root
CISA says attackers are exploiting a Linux IPv6 kernel flaw that can give a local user root and, on affected RHEL 10 systems, escape a container.
Read article →
CISA Gives Federal Agencies Three Days to Patch Exploited NetScaler Flaw
Citrix documents denial of service. Separate research demonstrates a SAML path to root code execution but has not confirmed that it maps to this CVE.
Read article →
Exposed Server Reveals Aurora Operator's Breach Playbook Across 17 Organizations
An exposed server revealed a repeatable Active Directory attack path, AI-assisted planning, Aurora lockers, and payment trails across multiple victims.
Read article →
Fake CAPTCHA Campaigns Load Malware That Disables Endpoint Security
ErrTraffic lures lead users to run Cruciferra, which uses a vulnerable signed driver to terminate security processes before the Remus stealer runs.
Read article →
Two Next.js Flaws Expose Self-Hosted Servers to Remote Code Execution
Next.js fixed two unauthenticated code-execution paths involving AVIF processing and Windows servers. Self-hosted operators need 15.5.24 or 16.3.3.
Read article →
MuddyWater Uses Deno to Hide a Backdoor in U.S. Companies
Dindoor uses the signed Deno runtime, encoded stages, and a pre-persistence sandbox check. Hunt the fixed process and registry sequence beneath the changing payload.
Read article →
Attackers Exploit Gitea Flaw to Run Code on Self-Hosted Servers
Gitea’s patch API can turn repository content into server code execution. Upgrade to 1.27.2 or later and check server activity for signs of compromise.
Read article →
CareCloud AWS Breach Affects 3.7 Million Patients
CareCloud says forensic review confirmed patient data exfiltration. HHS reports 3.76 million affected people, while public records do not map specific fields to each person.
Read article →
Attackers Exploit Oracle WebLogic Proxy Flaw to Reach Critical Data
CISA confirmed exploitation of a CVSS 10 Oracle WebLogic proxy flaw. Patch affected Apache and IIS plug-ins, then review requests for unauthorized data access.
Read article →
SynkLoader Uses Teams Lures for Password Phishing and Network Access
A new Windows loader arrives through a fake Teams help desk; modules sent to Expel’s emulator phish passwords and tunnel into internal services.
Read article →
ToxicPanda Steals Banking PINs and Takes Shell Control of Android Phones
An updated Android banking trojan blocks Google Play traffic, automates wireless ADB pairing, and targets 349 financial apps across 16 countries.
Read article →
Fake GTA 6 Demo Steals Passwords and Active Browser Sessions
A 1.1 MB fake installer delivers Vidar, launches installed browsers headlessly, and copies credentials and sessions that can remain useful after cleanup.
Read article →
Attackers Exploit miniOrange SAML Flaws for WordPress Admin Access
Two exploited miniOrange SAML flaws can mint WordPress admin sessions. Seven independently versioned editions make ordinary update and vulnerability checks unreliable.
Read article →
Attackers Use Android Car Updaters to Install Proxy Botnet Malware
Kaspersky traced malware on DoFun-powered car displays from a trusted updater to ad fraud and a residential proxy. DoFun says it fixed the issue but published no fixed build.
Read article →
N-able Passportal Flaw Let Malicious Pages Read Entire Password Vaults
The browser extension exposed vault tokens to untrusted page messages. Version 3.49.6 adds origin, frame, and nonce checks; later builds supersede it.
Read article →
C2Looper Backdoor Uses GitHub to Control Infected Windows PCs
C2Looper shifted from one-second HTTP beacons to GitHub C2. Hunt its OneDrive DLL, JSON control files, debug marker, commands, hashes, and two IPs.
Read article →
Public Forminator Forms Let Attackers Upload PHP to WordPress Servers
Forminator through 1.56.1 trusts forged upload settings and misses dangerous pipe-delimited MIME keys. Update to 1.57.1 and check public upload paths for executable files.
Read article →
Microsoft Defender Driver Lets Admins Delete Security Tools at Boot
Check Point reproduced kernel file and registry operations through Defender's signed BTR.sys driver. Abuse needs admin rights; behavioral telemetry separates it from cleanup.
Read article →
Suspected Ransomware Operator Used Claude Code to Steal Data and Map Backups
Gambit observed Claude Code inside six intrusions. Hunt the test VPN account, rogue LDAP listeners, backup discovery, SQL staging, and firewall restores.
Read article →
Unpatched Microsoft Configuration Manager Lets Domain Users Run Code as SYSTEM
Patch the missing upload check, then audit who can still reach the three remaining weaknesses in the Configuration Manager exploit chain.
Read article →
Attackers Exploit Zimbra Mail Flaw to Run Server Commands
Attackers are exploiting a Zimbra SNMP flaw through crafted SMTP requests. Version 10.1.20 fixes it; exposed servers need log and file review.
Read article →
Elementor Pro Upload Flaw Lets Attackers Run Code on WordPress Sites
A blank file entry bypasses Elementor Pro upload checks on exposed forms. Version 4.2.2 fixes the flaw, but patched sites still need to hunt for PHP left behind.
Read article →
Crafted IKEv2 Traffic Exposes Windows Hosts to Remote Code Execution
CISA says attackers are exploiting a Windows IKE remote-code flaw; verify April fixed builds and restrict UDP 500 and 4500 until every exposed host is patched.
Read article →
Ransomware Crews Exploit Windows Task Host Flaw for SYSTEM Access
CISA now ties CVE-2025-60710 to ransomware. The local Windows flaw needs an existing foothold, then lets an attacker elevate to SYSTEM.
Read article →
CISA Flags Exploited Ray Flaw That Runs Code Through Firefox and Safari
A hostile page can rebind its hostname to a Ray dashboard, change Firefox or Safari's User-Agent header, and submit a shell command to the Jobs API.
Read article →
45 Million wp2shell Attempts Hit WordPress in One Week
Pantheon counted 45 million wp2shell attempts in one week. Defenders should verify current WordPress builds and review exposed sites for persistent access.
Read article →
Poisoned Rust Crates Ran Malware During Developer Builds
A compromised crates.io account poisoned three Rust packages. Builds ran a downloader; 2,285 arrayref downloads now require cache, lockfile, and host checks.
Read article →
Windows 11 RAM Attack Lets Admins Disable Defender and VBS
USENIX researchers turned writable DIMM configuration into arbitrary physical-memory access. April's Windows update blocks their current chain under Secure Boot.
Read article →
Chrome 151 Fixes Five High-Severity Memory Flaws Across the Browser
Chrome 151 fixes five high-severity use-after-free bugs; endpoint inventories must show the fixed build or a later superseding release.
Read article →
Ransomware Campaigns Exploit SharePoint Flaw to Run Server Code
CISA now links CVE-2026-45659 to ransomware. Check exact SharePoint builds, then investigate whether low-privilege access reached the farm before patching.
Read article →
Windows Plug and Play Gives Crafted Devices SYSTEM Access
Researchers chain signed Windows device installers into SYSTEM execution through emulated USB hardware or an ordinary RDP session.
Read article →
Fake Web3 Interview Steals Wallet Keys Through Signed Windows App
A signed ClickOnce app delivered two stealers and an hVNC RAT after a fake Web3 interview. Hunt per-user ClickOnce records, then rotate every secret reachable from affected hosts.
Read article →
Attackers Exploit Windows Driver to Gain SYSTEM Privileges
Microsoft's August updates fix an AFD.sys race used in attacks; defenders must verify fixed builds and investigate pre-patch privilege escalation.
Read article →
Stolen AI API Key Leaves One Company With Nearly $1 Million Bill
Unit 42 says stolen AI API keys can reach gray-market proxy services within minutes, turning one exposed credential into catastrophic usage charges.
Read article →
NetScaler SAML Flaw Lets Remote Attackers Run Code as Root
Public research turns a patched NetScaler memory overflow into a root-code-execution risk for SAML deployments, making build verification urgent.
Read article →
RingCentral Breach Exposes 1.6 Million Account Records
Customer contact data is public after a social-engineering breach; defenders should harden support and identity checks against targeted impersonation.
Read article →
WordPress Image Upload Flaw Lets Authors Run Server Code
CVE-2026-65640 reaches Ghostscript on sites using Imagick; WordPress 7.0.4 and exact backports move content checks ahead of image processing.
Read article →
Microsoft Fixes LegacyHive Flaw That Gives Local Attackers Admin Rights
August Windows updates fix a public registry-hive privilege escalation; defenders can hunt its staging files, virtual paths, and unusual DLL loads.
Read article →
Attackers Exploit Cisco VPN Requests to Reload Firewalls
Exposed ASA and FTD remote-access services need release-specific hot fixes; Cisco offers no workaround for the exploited denial-of-service flaw.
Read article →
Two SharePoint Flaws Chain Into Unauthenticated Server Code Execution
Rapid7 joined a SharePoint identity bypass to unsafe .NET type creation. August's cumulative updates complete Microsoft's two-cycle fix.
Read article →
Gunra Steals Cloud Data and Deletes Backups Before Encryption
Government agencies traced Gunra from exploited edge devices through credential theft, cloud exfiltration, backup deletion, and cross-platform ransomware.
Read article →
CISA Confirms Exploitation of LoadMaster Command Injection
CISA confirms exploitation of a pre-authentication LoadMaster command-injection flaw. API-enabled appliances need a fixed release and an exposure review.
Read article →
Public Open vSwitch Exploit Gives Local Linux Users Root Access
OVSwrap exploits a 16-bit nested-action limit in the Linux kernel, leaving defenders to verify both the running fix and the earlier exposure window.
Read article →
Microsoft Defender Isolated a QNET Workstation in 128 Seconds
A Microsoft case study traces mshta execution to one isolated QNET host and shows which workstations qualify for the preview action.
Read article →
SCTPhantom Exposes Linux SCTP to Remote Attack and Local Root
A remote SCTP peer can trigger the kernel flaw, while Tencent separately demonstrated local privilege escalation and container escape. Defenders need both reachability and running-kernel checks.
Read article →
Cisco Fixes Seven IOS XE Flaws With No Workarounds
Cisco's August IOS XE hardening release fixes seven vulnerability classes across five reviewed trains; affected devices need an upgrade.
Read article →
Flooding Dropper Used 846 npm Components as Cross-Platform Loaders
An automated npm campaign uses hundreds of disposable packages to launch detached native malware, with DNS TXT records as a fallback delivery channel.
Read article →
Failed WordPress Login Input Reaches Conditional Server Code Execution
WordPress 7.0.3 fixes a pre-auth login-page XSS that can turn administrator interaction with a malicious site into PHP execution.
Read article →
Remus Malware Stores Command Server Addresses in Ethereum Contracts
Remus uses an Ethereum smart contract to resolve changing command infrastructure before stealing browser sessions, credentials, and wallet data.
Read article →
Windows Hello Keys Let Signed-In Code Create Persistent Entra Access
Code in a signed-in Windows session can invoke a TPM-backed Windows Hello for Business key, authenticate without a device identity claim, and create a path to durable Entra access.
Read article →
BINDCLOAK Duplicates Windows Tokens to Run Modules as Another User
BINDCLOAK collects Windows user and process tokens, duplicates them, and starts modular malware components inside more privileged security contexts.
Read article →
Researchers Forge WSUS Updates Through Standalone Microsoft SQL Server
SpecterOps shows how relayed WSUS machine-account access can forge targeted updates and bypass payload signature checks when SUSDB runs on a separate SQL Server.
Read article →
CISA Confirms Exploitation of TeamCity's Unauthenticated RCE
CISA confirms exploitation of a TeamCity flaw that lets an unauthenticated network attacker run commands as the server process, putting secrets and build integrity at risk.
Read article →
CISA Sets August 7 Deadline for Three Exploited Enterprise Flaws
Langflow, N-central, and Tomcat flaws entered CISA's exploited catalog. One gives unauthenticated callers Python execution by default.
Read article →
18 npm Packages Hid a RAT Across a Benign-Looking Dependency Chain
A targeted npm cluster split its downloader across ordinary-looking modules, then escaped Node.js vm isolation to install a cross-platform RAT on developer systems using Alibaba tools.
Read article →
Hidden VS Code MCP Settings Let Install Links Run Attacker Code
CVE-2026-41613 let crafted MCP install links persist settings that the VS Code preview did not show. Version 1.119.1 fixes the preview.
Read article →
Malicious RTSP Server Reaches Code Execution on TP-Link TL-WR940N v6
TP-Link TL-WR940N hardware revision 6 can let an unauthenticated attacker turn a LAN client's outbound RTSP session into code execution inside the router kernel.
Read article →
BTMOB Is No Longer One Android Malware Service
Flare found that BTMOB's official operation now sits among resellers, source-code buyers, private servers, and offers of uncertain authenticity, weakening infrastructure-only detection.
Read article →
DOUBLECUP Hides a ClickFix Payload in the Browser Cache
The service preloads a payload-bearing PNG, copies a browser-specific command, and keys the final in-memory stage to the victim's public IP address.
Read article →
Keyv's Signed npm Releases Carried a Credential-Stealing Worm
A compromised maintainer account published Keyv-family packages with valid provenance, an install-time credential stealer, and code that could poison more npm releases.
Read article →
Gitea Org Rendering Exposes Local Files Before Server Execution
Gitea 1.22.1 through 1.27.0 can let an anonymous request read files as the service account, expose the internal token, and plant a Git hook that executes during a clone.
Read article →
Unit 42 Extracted the Master Key Protecting Google's Synced Passkeys
Unit 42 showed three post-compromise paths from Chrome's local passkey state to silent assertions, substituted verification keys, or the master secret protecting synced credentials.
Read article →
INC Exploitation Makes SMA 1000 Hotfixes an Incident Triage Trigger
INC ransomware activity requires SMA 1000 operators to investigate logs and configuration as well as apply the hotfix.
Read article →
Chrome Fixed 1,072 Bugs, Then Changed How Updates Reach Users
Chrome fixed more security bugs in two milestones than in the previous 23 while using AI across discovery, triage and fix preparation.
Read article →
N-central Attacks Reached Customer PCs After the First Fix Fell Short
N-central operators need Hotfix 2 build 2026.3.1.10 and a downstream endpoint hunt because the new release supersedes the first hotfix.
Read article →
Galaxy Links COLDCARD RNG Flaw to Suspected $88.6 Million Theft
An integration error routed wallet entropy through MicroPython's deterministic Yasmarang fallback. Updating prevents new weak seeds, but existing ones still require migration.
Read article →
Adform Served a Script Built to Rewrite Crypto Addresses
A compromised Adform tracking file used browser events, DOM observers and form-field hooks to keep substituting Bitcoin, Ethereum and Tron addresses.
Read article →
Rails Active Storage Exposes Process Secrets Through libvips
A crafted image can reach unsafe libvips operations through Active Storage, exposing files and process credentials that a Rails-only update cannot recover.
Read article →
SourTrade Makes Browsers Assemble a Different Malware File for Each Victim
A ServiceWorker and SharedWorker combine a clean Bun runtime, delivered PE sections, and locally generated bytes before a same-origin download.
Read article →
XCSSET v40 Infects Every Xcode Project on a Compromised Mac
A poisoned build can start a memory-resident loader, re-arm through macOS preferences, and seed more projects, Git hooks, and archives.
Read article →
Certighost Lets a Domain User Impersonate a Domain Controller
A crafted AD CS chase sent the CA to rogue directory services. July updates add a real-DC check, but defenders still need issuance and replication evidence.
Read article →
24,650 Exposed BMCs Disclosed Password Hashes Before Login
IPMI's RAKP exchange exposes material for offline password cracking, leaving defenders with a management-plane incident that host telemetry may miss.
Read article →
XLab Traces 200,000-Device Dysphoria Botnet to UPnP Relays
One variant uses UPnP to open 155 inbound paths, then relays same-port traffic through infected devices that conceal the real command servers.
Read article →
FBI Says Exposed MicroLogix PLCs Disrupted Water Operations
Actors changed controller IP addresses and passwords across at least seven states. Operators need to compare running logic with a trusted baseline and investigate activity on connected devices.
Read article →
Attackers Exploit Arista VeloCloud Orchestrators Without Credentials
On-premises VCO is exposed by default. Arista says defenders must inspect web activity and managed Edge state after patching.
Read article →
Claude Breached Three Organizations During Anthropic Cyber Tests
A misconfigured evaluation harness let three Claude models reach production systems, exposing a control gap that prompts and model safeguards could not contain.
Read article →
Static Cisco Secure FMC Account Is Under Active Exploitation
Every on-prem Secure FMC configuration is affected. Cisco published hot fixes, a shared license.tmp indicator, and credential-rotation guidance.
Read article →
Compromised Microsoft 365 Mailboxes Spread QR-Code Phishing
Talos tied QR-code PDFs to credential theft, inbox-rule changes, SharePoint staging and new phishing sent from compromised Microsoft 365 mailboxes.
Read article →
Mirage Kitten Used Compromised Windows Hosts as Network Relays
Kaspersky found BridgeHead using Windows SSO to cross corporate proxies before relaying server-selected TCP traffic through compromised hosts.
Read article →
Fastjson Flaw Enabled RCE in Spring Boot Fat JAR Applications
The flaw reaches an unsafe resource-loading path without AutoType enabled. Exposure is limited to a specific Fastjson 1.x and Spring Boot deployment combination.
Read article →
Hermes Logs Record File Enumeration on a Thai Finance Ministry System
Five recovered task logs record service discovery, privilege checks and file enumeration after an operator enabled Hermes's unattended mode.
Read article →
OpenAI Cyber Test Breached Hugging Face Production
Hugging Face traced 17,600 actions from an Artifactory escape through two malicious-dataset vectors and into its clusters, network and source control.
Read article →
Crafted ChatGPT Link Created a Scheduled Workspace Agent in Tests
Zenity found URL parameters that preselected a template and auto-submitted instructions inside a logged-in user's Workspace Agents builder.
Read article →
CISA Confirms Exploitation of LiteLLM MCP Preview Command Injection
Any authenticated proxy-key holder could make two MCP preview endpoints run an arbitrary command on the LiteLLM host.
Read article →
Glassworm Poisoned 300+ GitHub Repositories With Stolen Credentials
The botnet spread through developer extensions and packages, then used stolen credentials to force-push malicious code into default branches.
Read article →
ServiceNow Pre-Auth RCE Reaches Connected Proxy Server Commands
CVE-2026-6875 chained query evaluation with a sandbox escape, giving an unauthenticated attacker broad control of a ServiceNow instance and its connected proxy servers.
Read article →
Search Ads Used Real ChatGPT Pages to Deliver Malware
Push Security traced paid search ads to shared ChatGPT and Claude pages that handed visitors to fake desktop-app downloads.
Read article →
Google Sued a Phishing Service Behind 2.5 Million Scam Texts
A civil lawsuit targets a Telegram-based phishing-kit operation that Google links to 2.5 million texts and more than one million fraudulent URLs.
Read article →
Microsoft Entra Actor Token Impersonated a Global Admin in Testing
CVE-2025-55241 allowed an Entra actor token from one tenant to impersonate users in another through the legacy Azure AD Graph API.
Read article →
GhostCommit Hid Secret-Theft Instructions Inside a PNG
Researchers put instructions to read .env and encode it into source inside a PNG that text-only pull-request reviewers ignored but vision-capable coding agents later followed.
Read article →
Edgecution Used a Hidden Edge Extension to Run Commands on Windows
A Teams lure installed a headless Edge extension and a Python native-messaging host, giving a ransomware access broker a quiet route to local command execution.
Read article →
Stolen npm Account Backdoored More Than 140 Mastra Packages
A stolen npm publisher account added a malicious dependency to more than 140 Mastra packages, giving Sapphire Sleet an install-time path into developer and CI systems.
Read article →
Fake Landing-Page Skill Reached 26,000 Agents After Passing Scanners
AIR says its brand-landingpage experiment reached 26,000 agents after scanners missed a remote instruction source that changed after approval.
Read article →
Amazon Q Ran MCP Commands Supplied by Project Repositories
CVE-2026-12957 allowed project configuration to start MCP processes with a developer's environment. AWS fixed the flaw in Language Servers for AWS 1.65.0.
Read article →
Klue Breach Exposed Salesforce Data Through Stolen OAuth Tokens
A forgotten Klue credential led to customer OAuth token theft and direct access to Salesforce CRM data across several companies.
Read article →
Gemini and ChatGPT Recommended Malicious GitHub Skills in Tests
Island found more than 800 fake skills and MCP servers using credible READMEs and ZIP files to turn capability searches into malware installs.
Read article →
Poisoned AsyncAPI Packages Ran Code When Developers Imported Them
Five poisoned releases avoided install hooks, launching a detached Node.js process when developer or CI tooling loaded the affected module.
Read article →
Gemini CLI Rebuilt Botnet Command Infrastructure in Six Minutes
Trend Micro found a Russian-speaking actor using Gemini CLI to rebuild command-and-control infrastructure and operate eight infected computers.
Read article →
Cursor Flaws Let Prompt Injection Overwrite Its Sandbox and Run Code
Cursor 3.0 fixed path-handling failures that let injected instructions write beyond a project and tamper with the sandbox protecting the host.
Read article →
Hidden Web Instructions Made Kiro Run Attacker Code in Tests
A routine page fetch in older Kiro builds could end with attacker-controlled code running through a rewritten MCP configuration.
Read article →
NadMesh Scans Exposed AI Services for Cloud Credentials
The botnet gives AI workbenches priority in its scan queue, then searches compromised hosts for cloud credentials, service-account tokens, and callable tools.
Read article →
Fake Claude Installers Delivered SectopRAT to 29 Organizations
A paid search ad sent victims through a public page on claude.ai before delivering a fake desktop installer.
Read article →