Blog & Resources
Security awareness tips, phishing prevention guides, and best practices.
Nvidia-led OSAA grows to 120+ firms, forms Shared AI
The Open Secure AI Alliance, launched by Nvidia, now includes over 120 companies and has created the Shared AI Findings working group to tackle AI sec
AI guardrails easily bypassed by simple claims, Cisco Talos
Cisco Talos researchers found that AI guardrails in tools like Claude Code and Codex are easily bypassed by simple claims of ownership or participatio
Hacker Summer Camp 2025: AI Dominates Black Hat, DEF CON
As the security industry gathers in Las Vegas for BSides, Black Hat, and DEF CON, agentic AI security and governance take center stage, with keynotes
Microsoft Expands Zero Trust for AI with New Security Tools
Microsoft announced new tools and guidance to secure AI agents and DevSecOps environments, extending its Zero Trust for AI strategy.
cMCP Lets AI Agents Deny Calls with Signed Receipts
cMCP, launched June 23, 2026, enables AI agents to deny tool calls and issue signed receipts for accountability.
White House AI Tests Could Signal Trust for Banks
The White House is creating a voluntary pre-market testing checkpoint for advanced AI models.
AWS launches Kiro Crew autonomous AI coding orchestrator
AWS unveiled Kiro Crew, an autonomous workspace that runs AI coding agents around the clock, letting developers delegate tasks and step away.
AI Lowers Hacking Barrier, Enabling Novice Attackers
AI tools are enabling inexperienced attackers to perform sophisticated hacking, challenging the assumption that offensive capability requires deep tec
Obsidian Security raises $85M for AI agent security
Obsidian Security secured $85M in Series D funding at a $1.1B valuation to protect cloud apps from AI agent threats.
Google Removes 3 ADK AI Workflows After Prompt Injection
Google deleted three AI agent workflows from its Agent Development Kit after Pillar Security found a public GitHub issue could prompt-inject a triage
Mistral AI Unveils Shieldstral 3B Safety Model
Mistral AI released Shieldstral, a 3B open-weights multimodal moderation model that matches text safety of models 7x larger, setting a new state of th
AI Security Needs Interaction-Aware Layer Beyond CASB and
Traditional CASB and DLP tools are insufficient for securing AI use. Organizations need an interaction-aware layer to monitor and control employee-AI