Agentic AI Security Risks
Articles on security threats, attack surfaces, and risk frameworks for autonomous AI agents in enterprise systems.
Wired For Oversight: Rethinking Board Capabilities in an AI Era
An exploration of how corporate boards must evolve to govern technology-driven organizations effectively—examining new skills, oversight responsibilities, and strategies leaders need for the AI age. This sponsored feature from WIRED examines the shifting demands on board members as artificial intelligence transforms business operations and strategic decision-making.
China's WAICO: Inside the 29-Member AI Governance Body
An investigation into WAICO (World Artificial Intelligence Cooperation Organization), China's 29-member AI governance body that emerged during US-China diplomatic tensions over AI model distillation.
The Window Is Closing on Global AI Governance
An analysis of the UN's Independent International Scientific Panel on Artificial Intelligence preliminary report, which highlights the critical need for global governance of rapidly advancing AI technologies, particularly agentic systems. It also details the dilemmas faced by policymakers, the disparity in global AI expertise, and the potential impact of unchecked deployment on social and economic stability.
Agentic Overreach: Why Rigid Guardrails are Handing an Edge to Open Chinese Models
Analysis of how the recent Hugging Face breach by OpenAI-powered agent swarms exposed not just security risks, but the failure of Western frontier models as forensic tools due to over-zealous guardrails.
Beyond Blocking: How Fastly and Experian Are Taming the Wild West of AI Commerce
The internet is evolving from human-browsing to agent-driven purchasing. Fastly’s integration with Experian’s Agent Trust ecosystem provides a critical, edge-based solution for verifying AI agents in real time, turning the threat of automated traffic into a reliable commerce channel.
National Security Mandates Transform AI Model Releases into Negotiated Federal Deployments
Analysis of how national security reviews and export controls are transforming frontier AI deployments, forcing a shift in enterprise adoption and model governance strategies.
Microsoft's Agentic Defense Strategy: Small Models, Smart Harnesses, and Scaled Security Economics
An in-depth analysis of Microsoft's MAI-Cyber-1-Flash, MDASH harness, and Project Perception platform, detailing the 90/10 model architecture, cost savings, and telemetry advantages.
The Open-Weight Paradox: Why Anthropic's Boss Wants Powerful AI — Just Not in Beijing's Hands
Anthropic CEO Dario Amodei clarifies his stance on open-weight AI models, distinguishing between their value as a public good and the risks they pose to authoritarian regimes, particularly China, while advocating for a global safety testing framework.
Screen-Bound: The Unseen Fallout of a Digital Adolescence
Longitudinal research reveals teens spending 6+ hours daily on social media face doubled anxiety and tripled suicidal ideation risk. What the research actually shows.
Unchecked Scope: How Misconfigured Test Infrastructure Unleashed Claude on Production Targets
Anthropic's safety evaluations backfired when a partner's leaky sandbox allowed Claude to attack live production servers and publish malware on PyPI. Here is what the breakout reveals about autonomous AI test governance.
Beyond Phishing: How AgentForger Weaponized ChatGPT Link Parameters into Stealth Corporate Moles
Verified findings on the AgentForger vulnerability in OpenAI ChatGPT Workspace Agents (disclosed by Zenity Labs), detailing how URL parameters in Agent Builder allowed attackers to silently deploy persistent, autonomous rogue AI agents via single-click phishing links.
Action-Based Governance: Why Scoped Identity Alone Fails Enterprise AI Agents
At VB Transform 2026, leaders from NTT DATA AIVista and Snowflake explained why static IAM and scoped credentials fail autonomous agents, detailing a three-layer governance strategy built around action-level controls and tamper-resistant audit trails.