**New “BioShocking” Attack Exposes Flaw in Agentic Browsers** A group of researchers has discovered a novel attack technique that exploits a weakness in several popular agentic browsers, allowing attackers to manipulate these AI-powered tools into stealing sensitive credentials. The attack, dubbed “BioShocking,” was demonstrated by LayerX, a cybersecurity firm, and could have far-reaching implications for … Read more
Cisco Warns of In-the-Wild Exploitation of Unified Communications Vulnerability A critical vulnerability in Cisco’s Unified Communications Manager (Unified CM) and Unified Communications Manager Session Management Edition (Unified CM SME) has been exploited in real-world attacks, according to a warning issued by the company. The flaw, tracked as CVE-2026-20230 with a CVSS score of 8.6, allows … Read more
The Trump Administration Lifts Restrictions on Anthropic’s AI Models After Cybersecurity Alarm In a significant development, the Trump administration has lifted restrictions on artificial intelligence company Anthropic’s latest versions of its Claude chatbot, following weeks of concerns over potential cybersecurity risks. The decision comes as part of a broader effort to vet national security threats … Read more
Cursor AI Code Editor Flaws Expose Users to OS-Level Remote Code Execution A pair of critical vulnerabilities in Cursor, a popular artificial intelligence code editor, has been discovered by security researchers at Cato Networks. The flaws, designated as CVE-2026-50548 and CVE-2026-50549, have a CVSS score of 9.8, indicating a high risk of exploitation. These issues … Read more
Google, FBI Disrupt Massive Residential Proxy Network Powered by Millions of Devices In a major blow to cybercrime operations, Google has joined forces with the FBI and other organizations to dismantle NetNut, a massive residential proxy network that had infected over 2 million Android devices. The takedown is a significant development in the ongoing battle … Read more
A large-scale credential-harvesting operation targeting over 430,000 FortiGate firewalls worldwide has led to the deployment of ransomware against hundreds of organizations. The campaign, dubbed FortiBleed, has been linked to two prominent ransomware families: INC Ransom and Lynx. Cybersecurity firm SOCRadar reports that the attackers have compromised over 110 million credentials since February, with a significant … Read more
**Cybersecurity Experts Warn of Hidden Risks in AI-Driven Software Development** A growing number of organizations are leveraging artificial intelligence (AI) and large language model (LLM)-assisted code to boost productivity and efficiency in software development. However, a recent report reveals that one in five companies has experienced a serious security incident directly tied to AI-generated code. … Read more
A new CitrixBleed-like vulnerability has been exploited by threat actors barely 24 hours after its public disclosure. The bug, tracked as CVE-2026-8451 with a CVSS score of 8.8, affects NetScaler ADC and NetScaler Gateways configured as SAML IDP, allowing attackers to access sensitive information without authentication. The vulnerability was discovered in the XML parser of … Read more
Claude Fable’s Relaunch Leaves Users Underwhelmed by Subpar Performance The long-awaited relaunch of Claude Fable, a powerful AI model from Anthropic, has been met with disappointment and frustration among users. Despite being made available to all users, including those with lower-tier subscriptions, the model’s performance falls short of expectations. According to reports, Fable 5 is … Read more
A popular AI model, Fable 5, is temporarily leaving subscription plans due to high demand, but its creators at Anthropic assure users it’s not a permanent change. The news has sparked concerns among users who rely on this powerful language model for their work. For those unfamiliar with the situation, Fable 5 was recently restored … Read more