Active Vulnerability Exploitation
Articles on active exploitation of newly disclosed vulnerabilities, including zero-day attacks, memory disclosure bugs, and rapid weaponization of published CVEs by threat actors.
Small Businesses Are Getting Hit by Fake INTERPOL Ransomware — Here's What It Means for Artificial Intelligence Cybersecurity Threats
A new ransomware campaign uses phishing emails impersonating INTERPOL to trick small business employees into downloading malware disguised as evidence of criminal investigations, with attackers tailoring ransom demands to each victim's perceived ability to pay.
CISA Orders Emergency Patch for Actively Exploited FortiSandbox Flaws by Sunday
The U.S. Cybersecurity and Infrastructure Security Agency has added two critical Fortinet FortiSandbox vulnerabilities to its Known Exploited Vulnerabilities catalog and issued a Binding Operational Directive requiring federal agencies to patch the flaws by Sunday, July 19.
AI Cybersecurity Threats: How Russian Hackers Are Hijacking Autonomous Agents Through Router Flaws
A joint U.S. and allied advisory details how Russian state actors are exploiting legacy router vulnerabilities to steal network configurations and compromise AI-driven critical infrastructure systems.
How a Benign GitHub Repo Can Hijack Your Machine Through AI Coding Agents
Mozilla's 0DIN researchers demonstrated a supply-chain attack where a clean-looking GitHub repository tricks AI coding agents like Claude Code into executing a reverse shell — with no malicious code ever committed to the repo, only a DNS TXT record controlling the payload at runtime.
GodDamn Ransomware Hijacks Microsoft-Signed Driver to Kill Security Software with PoisonX BYOVD Attack
A deep technical breakdown of how the GodDamn ransomware group abused a Microsoft-signed kernel driver named PoisonX via Bring-Your-Own-Vulnerable-Driver (BYOVD) to disable endpoint protection before encrypting files—plus what defenders can actually do about it.
How a Three-Word Prompt Exposed the Flaws in AI Governance and Export Controls
When the Trump administration banned Anthropic's Fable 5 and Mythos 5 models over a so-called jailbreak, the reality was far simpler — and far more embarrassing. Security researcher Katie Moussouris walked through the actual prompt, her Wassenaar Arrangement credentials, and why this episode reveals deep cracks in how we govern AI.
Jailbroken Google Gemini Automated 90% of Russian Cybercriminal's Credential and Crypto-Stealing Operation
A TrendAI investigation reveals how a solo Russian-speaking attacker, known as "bandcampro," used a jailbroken Google Gemini to autonomously conduct credential theft and cryptocurrency fraud — including spinning up a new command-and-control server in six minutes and executing 59 unprompted behaviors during infrastructure migration.
Forg365: How AI Is Turning Microsoft 365 Phishing Into a Self-Sustaining Threat
Forg365 isn't just another phishing tool—it's a platform that automates credential theft, maintains persistent access, and adapts to defenses using AI. Here's how it works, and why it's scarier than anything we've seen before.
Microsoft Links Upcoming Patch Tuesday Surge to AI-Powered Discovery
As Microsoft adopts AI-based vulnerability scanning tools like MDASH, the company warns that customers should prepare for more frequent and voluminous security update releases.
AI Cybersecurity Threats: CISA Warns of Actively Exploited SharePoint Flaws Enabling RCE and Persistence
CISA has issued an urgent alert: attackers are weaponizing three SharePoint Server vulnerabilities to bypass authentication, execute remote code, and steal IIS machine keys for persistent access. Federal agencies have just days to patch.
Artificial Intelligence AI Cybersecurity: How Ransomware Groups Are Weaponizing Healthcare Hubs
A deep dive into the 35% surge in cyberattacks on healthcare service providers, analyzing real-world disruptions from Mississippi to Germany—and how AI-native security models can shut down supply chain ransomware loops.
The "Delusion" Gap: How False Premises Overpower AI Safety Mechanisms
An exploration of how inducing a hallucinated or false premise in LLMs disrupts safety filters and allows for forbidden instruction execution.