Druva announced Druva AI Resilience, a new approach for recovering, governing, and defending AI-powered work, introducing capabilities for Microsoft Copilot, Claude Code, Druva Model Context Protocol …
Scammers impersonating FBI personnel are targeting people who already lost money, using AI-generated video of FBI officials and spoofed versions of the IC3 website, the Internet Crime Complaint Center…
Amazon GuardDuty investigation agent, now in public preview, uses AI to automate initial threat investigation steps, reducing time for security teams. The feature is free during preview in 10 AWS Regi…
Most large companies now run multiple AI platforms simultaneously, with sanctioned tools and personal accounts coexisting inside organizations, according to anonymized sign-on data from more than 20,0…
GNOME is shortening its security disclosure window due to a surge in AI-generated vulnerability reports, according to Michael Catanzaro, who has run GNOME's security issue tracking since November 2020…
Attackers are actively exploiting CVE-2026-6875, a critical pre-authentication remote code execution vulnerability in the ServiceNow AI Platform, according to threat intelligence firm Defused. The fla…
Hugging Face disclosed a security breach carried out by an autonomous AI agent system that gained unauthorized access to internal datasets and credentials, the company said in a July 16 blog post. The…
Researchers have developed a forensic tool to detect backdoored code completions in AI coding assistants, addressing the risk of poisoned training data that can cause models to generate insecure code …
Nearly half of open-source AI projects never reach production, according to Mozilla's The State of Open Source AI 2026 report, which identifies deployment, governance and operational tooling as persis…
1Password has introduced 1Password for Claude, a beta integration that lets Anthropic's AI assistant complete browser tasks requiring authentication without accessing users' passwords or other secrets…
A group at UC Berkeley warns that prompt injection is becoming the cross-site scripting (XSS) of the web agent era, calling the vulnerability Cross-Site Prompting (XSP). Their system, Prismata, sits b…
A study by researchers at Harvard Kennedy School and Meta found that the script, not the voice, is what makes AI voice phishing effective. The team tested six commercial voice systems and human caller…
Polygraf AI released Meeting Guard, a real-time deepfake detection solution for enterprise meetings that joins as a visible participant and delivers near-real-time security analysis. The product was a…
A Russian-speaking threat actor known as “bandcampro” used a jailbroken Gemini CLI, Google’s open-source terminal-based AI agent, to deploy and operate a command-and-control botnet that controlled eig…
Intruder has launched AI Pentesting for web applications, enabling organizations to connect codebases via GitHub or GitLab to automatically scope and launch penetration tests in minutes, with results …
Tenable has expanded the Tenable One Exposure Management Platform to unify application security risks with enterprise exposure data, integrating static code vulnerability data for code-to-runtime visi…
Lineation.ai has publicly launched a comprehensive agentic security platform that combines GenAI Application Security with Runtime Defense, introducing a Zero Trust unified control plane and a lightwe…
Visual Studio Code's 1.129 release introduces a dedicated agent host process that runs AI coding agents such as Copilot, Claude, and Codex via the Agent Host Protocol, isolating each session in its ow…
Roughly 96 percent of codebases contain open-source software, according to a Help Net Security report, highlighting the critical dependence on volunteer-written code that underpins most corporate appl…
OpenAI's automated red-teaming model GPT-Red outperformed human red teamers on a prompt injection test, using self-play reinforcement learning to find weaknesses in GPT models. The attacker model earn…