Tag: AI security

Researchers have uncovered the FakeGit campaign using nearly 7,600 malicious GitHub repositories, including AI skills and MCP servers, to distribute SmartLoader and StealC malware.

Researchers have uncovered the NadMesh botnet targeting exposed AI services, cloud credentials, Kubernetes tokens, and administrative interfaces across internet facing systems.

Awan Distribution and TrendAI hosted their second Coffee Talk session in Islamabad, bringing together cybersecurity professionals to discuss ZTSA Internet Access, AI Secure Access, and emerging cyber threats.

Kaspersky has urged individuals and businesses to avoid sharing sensitive information with AI tools, as new research shows growing concerns about data security and privacy risks associated with artificial intelligence.

Researchers at Wiz have disclosed the GhostApproval vulnerability affecting several AI coding assistants, enabling malicious repositories to manipulate file approvals and potentially execute code on developer systems.

China's National Vulnerability Database has warned of alleged security backdoor risks in Anthropic's Claude Code, claiming the AI coding tool could transmit sensitive user information without consent.

This week's cybersecurity roundup covers the disruption of the NetNut proxy botnet, AI driven attack techniques, browser based ransomware, phishing campaigns, malware activity, and critical vulnerabilities affecting organizations worldwide.

This ThreatsDay roundup covers ransomware phishing campaigns, sandbox escapes, Apple email vulnerabilities, Linux kernel privilege escalation flaws, AI compute hijacking, and large scale malware operations affecting enterprise and consumer systems worldwide.

Check Point Research has identified AI generated browser ransomware that uses Chromium File System Access API to encrypt files on Windows and Android without requiring a native payload or browser exploit.

Adversa AI has disclosed GuardFall, a shell injection bypass affecting 10 of 11 tested open source AI coding agents, exposing systems to command execution and credential theft risks.

BeyondTrust outlines why identity security, privilege management, and assume breach strategies have become essential as enterprises face growing risks from AI driven attacks, machine identities, and trusted access abuse.

Google has patched a security flaw in the Vertex AI SDK for Python that could allow attackers to hijack machine learning model uploads through bucket squatting and execute malicious code within Google’s serving infrastructure.