Cybersecurity researchers have discovered a novel attack technique called TokenBreak that can be used to bypass a large language model's (LLM) safety and content moderation guardrails with just a single character change. "The TokenBreak attack targets a text classification model's tokenization strategy to induce false negatives, leaving end targets vulnerable to attacks that the implemented
AI is changing everything — from how we code, to how we sell, to how we secure. But while most conversations focus on what AI can do, this one focuses on what AI can break — if you’re not paying attention. Behind every AI agent, chatbot, or automation script lies a growing number of non-human […]
A novel attack technique named EchoLeak has been characterized as a "zero-click" artificial intelligence (AI) vulnerability that allows bad actors to exfiltrate sensitive data from Microsoft 365 Copilot's context sans any user interaction. The critical-rated vulnerability has been assigned the CVE identifier CVE-2025-32711 (CVSS score: 9.3). It requires no customer action and has been already
Human identities management and control is pretty well done with its set of dedicated tools, frameworks, and best practices. This is a very different world when it comes to Non-human identities also referred to as machine identities. GitGuardian’s end-to-end NHI security platform is here to close the gap. Enterprises are Losing Track of Their Machine […]
ConnectWise has disclosed that it's planning to rotate the digital code signing certificates used to sign ScreenConnect, ConnectWise Automate, and ConnectWise remote monitoring and management (RMM) executables due to security concerns. The company said it's doing so "due to concerns raised by a third-party researcher about how ScreenConnect handled certain configuration data in earlier versions.
Cybersecurity researchers have uncovered a new account takeover (ATO) campaign that leverages an open-source penetration testing framework called TeamFiltration to breach Microsoft Entra ID (formerly Azure Active Directory) user accounts. The activity, codenamed UNK_SneakyStrike by Proofpoint, has targeted over 80,000 user accounts across hundreds of organizations' cloud tenants since a surge in
Former members tied to the Black Basta ransomware operation have been observed sticking to their tried-and-tested approach of email bombing and Microsoft Teams phishing to establish persistent access to target networks. "Recently, attackers have introduced Python script execution alongside these techniques, using cURL requests to fetch and deploy malicious payloads," ReliaQuest said in a report
Threat intelligence firm GreyNoise has warned of a "coordinated brute-force activity" targeting Apache Tomcat Manager interfaces. The company said it observed a surge in brute-force and login attempts on June 5, 2025, an indication that they could be deliberate efforts to "identify and access exposed Tomcat services at scale." To that end, 295 unique IP […]
INTERPOL on Wednesday announced the dismantling of more than 20,000 malicious IP addresses or domains that have been linked to 69 information-stealing malware variants. The joint action, codenamed Operation Secure, took place between January and April 2025, and involved law enforcement agencies from 26 countries to identify servers, map physical networks, and execute targeted takedowns. […]
In today’s cybersecurity landscape, much of the focus is placed on firewalls, antivirus software, and endpoint detection. While these tools are essential, one critical layer often goes overlooked: the Domain Name System (DNS). As the starting point of nearly every online interaction, DNS is not only foundational - it’s increasingly a target. When left unsecured, […]