Patch now
06 / 28
▾
CISA warns of hackers exploiting critical MLflow vulnerability
CISA warns threat actors are actively exploiting a critical MLflow vulnerability to steal cloud credentials from federal and other systems.
Critical Zimbra RCE flaw now actively exploited in attacks
A critical RCE flaw in Zimbra Collaboration Suite is now being actively exploited in the wild, per CERT Polska warning.
ICS Operators Warned of AI-Driven Attacks on Siemens PLCs
NSA/CISA advisory warns ICS operators of attackers using AI-generated exploitation scripts against exposed Siemens S7 PLCs in critical sectors.
MLflow Vulnerability Exploited for Cloud Credential Theft
Actively exploited critical MLflow flaw allows SSRF to internal endpoints, enabling cloud credential theft; CISA has issued a warning.
Critical GitLab Flaw Exploited Shortly After Disclosure
CVE-2026-19478, a critical unauthenticated GitLab flaw allowing project/data modification, is already being exploited after disclosure.
Hackers Using AI to Target Siemens PLCs in Critical US Sectors
NSA, CISA advisory details AI-assisted attacks targeting Siemens S7 PLCs in critical US infrastructure sectors.
Monitor — breaches & campaigns
13 / 28
▾
NASA AIT-GUI Flaws Could Let Unauthenticated Attackers Issue Spacecraft Commands
Unauthenticated attackers can issue arbitrary spacecraft commands via chained flaws in NASA's open-source AIT-GUI toolkit, posing serious mission integrity risks.
ToxicPanda 2.0 and GoldDigger Expand Android Banking Attacks with On-Device Fraud
ToxicPanda 2.0 Android banking trojan adds 167 remote commands and expands global targeting, enabling on-device fraud across 140+ banking and crypto apps.
40 Malicious Firefox Extensions Pose as Web3 Products to Steal Wallet Secrets
40 malicious Firefox extensions impersonate popular Web3 wallets to steal cryptocurrency wallet secrets from users.
Elementor Pro Flaw Could Let Unauthenticated Attackers Upload PHP and Execute Code
Critical CVE-2026-32475 (CVSS 9.0) in Elementor Pro WordPress plugin allows unauthenticated PHP upload and remote code execution.
Citrix urges admins to patch new NetScaler flaws as soon as possible
Citrix urges immediate patching of two vulnerabilities in NetScaler Gateway and NetScaler ADC; exploitation expected imminently.
New Manic Android malware can exfiltrate data through nearby devices
New Android malware 'Manic' targeting European users uses nearby infected devices as a fallback exfiltration channel, complicating detection.
Updated ToxicPanda Variant Targets 140+ Banking and Crypto Apps
Duplicate coverage of ToxicPanda 2.0 Android banking trojan targeting 140+ apps; see item 2.
French tax authority says break-in exposed data of 600K, including some private messages
French tax authority breach exposed personal and financial data of 600,000 individuals, including private messages.
Atlassian, Splunk Patch Dozens of Critical, High-Severity Vulnerabilities
Atlassian and Splunk patched dozens of critical and high-severity flaws enabling RCE, privilege escalation, and sensitive data access.
Cisco Patches Critical Crosswork, Secure Workload Vulnerabilities
Cisco patched critical vulnerabilities in Crosswork and Secure Workload products covering RCE, authentication bypass, and path traversal.
Exploitation Expected for Critical Authentication Bypass Patched in Citrix NetScaler
Critical authentication bypass in Citrix NetScaler exploitable by unauthenticated remote attackers; exploitation widely anticipated.
UAT-10147 deploys SPECTRE: A cross-platform implant with Linux rootkit and BYOVD capabilities
Cisco Talos details SPECTRE, a new cross-platform implant with Linux rootkit, BYOVD kernel bypass, and credential theft used by UAT-10147.
UAT-10147: Chinese-speaking adversary integrates agentic AI into post-compromise operations
Chinese-speaking threat group UAT-10147 integrates agentic AI into post-compromise operations, targeting vulnerable web servers globally.
Watch — medium severity
08 / 28
▾
NCSC Urges Stronger Controls for Agentic AI Systems
NCSC recommends sandboxing, oversight, and tight access controls for autonomous AI agents to reduce security risks.
US Defense Contractors Admit Their Rising CMMC Scores May Not Be Accurate
US defense contractors admit their own CMMC self-assessment scores may be inaccurate, raising concerns about cybersecurity compliance integrity.
Def Con Attendees Targeted by Persistent Phishing Campaign
A Huntress researcher describes an elaborate, persistent phishing campaign targeting Def Con attendees post-conference.
Grok chat duped into swallowing injected instructions
Grok AI chatbot was manipulated via prompt injection using encrypted instructions to bypass content controls and spread malware.
AI agent suggested installing a malware package. Engineer almost took its advice
An AI coding agent recommended installing a malicious package; the incident highlights supply-chain risks from AI-generated code suggestions.
AI-Assisted Tool Helped Secure Satellite Communication System After 2022 Russian Hacking
AI-assisted tool Argo is being used to verify and harden Viasat's satellite network resilience following the 2022 Russian cyberattack.
OpenAI Overhauls Model Security With Sandboxing, 30-Minute Alerts, and Training Pauses
OpenAI introduces sandboxing, 30-minute security alerts, and training pauses to strengthen AI model security following recent incidents.
Identity Abuse Through Trusted Communication Channels
Unit 42 details how attackers abuse enterprise collaboration tools like Teams and Slack for identity phishing and credential theft.
Awareness — policy, research, signals
01 / 28
▾
Using Microsoft Graph and Powershell to Mine for Information - Stale Accounts and Licenses, (Thu, Aug 20th)
SANS guide on using Microsoft Graph API and PowerShell to identify stale accounts and licenses for security hygiene and attack surface reduction.