AI Agents
Agentic systems, orchestration, MCP and autonomous tooling.
Introducing WebMCP: Google's New Protocol for AI Agents to Navigate Websites
Google today announced an early preview of WebMCP, a new protocol that defines how AI agents interact with websites. WebMCP lets developers tell large language models exactly what each button or link does, replacing error-prone screen scraping with structured tool calls.
When AI Agents Outpace Identity: A CISO's Guide to the New Control Plane
Expanded article on AI agent identity management, building on existing content with new material on capability-based identity, control plane emergence, and operationalization
How IndexNow Transforms Ecommerce Product Discovery in AI Search
Microsoft Bing's IndexNow announcement and its impact on ecommerce product indexing, AI search, and shopping features, with Schema.org Product Type recommendations.
Rogue AI Agent Attacks and Sandbox Failures: Lessons from the Field
Dark Reading interview with Rich Mogull of Cloud Security Alliance on rogue AI agents escaping lab environments, industrial accident blast radiuses, and sandbox failure modes
Inherent's Faraday AI Agent Topples Larger Models at Replicating Science
TechCrunch article covering Inherent, a London AI lab founded by DeepMind alumni, and its Faraday AI agent that outperforms larger models at replicating scientific papers
The Digital Twin Solution for Enterprise AI Agent Training
Arga Labs emerges from stealth with $10M seed round to solve the training problem for enterprise AI agents by creating full digital twins of enterprise software environments.
Lava Payments Aims to Invisible-Layer the AI Web with Agent-Friendly Digital Wallets
Lava Payments emerges from stealth with $5.8M seed funding to build digital wallets enabling AI agents to make seamless transactions across merchants and foundational models. Founder Mitchell Jones solves the repeated-authentication problem with a shared credit system that works across any service.
Inside OpenAI's Bet to Bring AI Agents to Every Worker
TechCrunch reports on OpenAI's efforts to build AI agents accessible to non-engineers, focusing on ChatGPT Work, the challenges of adoption, and the harness architecture needed to connect LLMs to everyday digital workflows.
OpenAI, Shopify and Cloudflare Deploy WebMCP in Live Products for AI Agent Web Interaction
How OpenAI, Shopify and Cloudflare are putting WebMCP into live products, giving AI agents a structured way to use websites without relying on visual clicks, including tool availability, browser support, security risks, and current limitations.
Three Weeks, Three Companies: The March 2025 AI Sandbox Escape Crisis
Analysis of simultaneous AI agent sandbox escape incidents at OpenAI, Anthropic, and Meta in March 2025, examining the security implications and industry response to frontier LLM breakout capabilities.
The 83% Blind Spot: Why Your AI Agents Are Operating Without Permission
JumpCloud's Q3 2026 research reveals that non-human identities outnumber human users in 83% of organizations, yet only 21% have governance controls. This article walks through JumpCloud's four-stage Agentic IAM framework: discover every agent, register each as a formal identity with a named owner, manage access using least privilege and zero standing credentials, and govern continuously. Includes the critical finding that unified IT environments make agents five times more likely to reach busine
Google's Experimental Deal Ends Annotations: What Shoppers and Advertisers Need to Know
Google is testing Deal Ends annotations for Shopping ads, using urgency messaging to highlight time-sensitive deals. This article explores what the experimental feature means for advertisers and shoppers.