Google, Anthropic, and OpenAI Unveil Cyber AI Models, Safeguards, and Access Programs

Google on Wednesday announced Gemini 3.8 Flash Cyber, which it described as its most capable cybersecurity model, and has made it available to a set of trusted defenders via a new initiative called the Fairwind Program. "The Fairwind Program gives high-priority defenders like governments,...

Wordfence Argus: Moving Beyond Human Research Capability

When you create an AI agent that makes a breakthrough that is so difficult to understand that you need to ask it to write a blog post to explain it to you, you know you're on to something. If you're paying attention, you should have noticed that the major vulnerability we recently reported in one...

PT-2026-81211

Name of the Vulnerable Software and Affected Versions U-Boot versions prior to 2026.10-rc5 Description A use-after-free issue exists in the lwIP wget implementation. In the httpc recv cb function, when HTTP data storage fails, the callback frees the connection PCB Protocol Control Block but retur...

Agent389

Agent389 Agent389 is a professional, high-fidelity LDAP inje...

exploitra

🚀 EXPLOITRA v1.0 - Advanced Vulnerability Engine EXPLOITRA...

Claude Mythos: Preparing for a World Where AI Finds and Exploits Vulnerabilities Faster Than Ever

Anthropic's new model can autonomously discover zero-days and develop working exploits. While access is currently limited to responsible actors, now is the time to strengthen response playbooks, reduce exposure, and incorporate AI into security programs...

How Agentic AI Helps with Adaptive Cloud Risk Assessment with Agent Vikram

In fast-moving cloud environments like AWS, security teams face an uncomfortable truth: not every EC2 instance is being scanned, existing tools don’t work across a diverse environment that includes long-lived and ephemeral assets, and visibility is never complete. Qualys research found that over...