Posts

“AI agent” is becoming the new “cloud-native.” Every software vendor, platform provider, and enterprise SaaS company is rushing to attach the label to their products. According to Gartner’s August 2025 forecast, 40% of enterprise applications will be integrated with task-specific AI agents by end of 2026 — up from less than 5% in 2025. Gartner has a name for what’s driving much of that growth: agentwashing. Agentwashing is the practice of relabeling automation, robotic process automation (RPA), or chatbots as “agentic AI” for marketing purposes. It’s not a fringe phenomenon. It’s happening across the enterprise software landscape, and it has real consequences for buyers making decisions based on inflated capability claims. ...

The question of how do you govern what an AI agent is actually doing at runtime has been one of the most urgent unsolved problems in production agentic AI. On May 27, 2026, at the AI Agent Security Summit in San Francisco, the Agent Control Standard (ACS) launched as the first open, vendor-neutral framework for runtime governance of AI agents. ACS is not another AI safety research paper. It’s a concrete technical specification — Apache 2.0 licensed, developed in the open via community consensus — for intercepting, controlling, and auditing AI agent actions in real time. ...

Something significant happened in the spring of 2026 and it happened fast: in a six-week window, Google, Anthropic, and AWS all shipped managed AI agent runtime infrastructure. The New Stack — corroborated by VentureBeat’s detailed analysis — is arguing that the runtime layer has effectively been commoditized. The competitive frontier for AI is moving up the stack. The convergence has real implications for teams building production agent systems today. The Six-Week Sprint The timeline of what happened: ...

One of the most persistent blockers for enterprise AI agent adoption just got removed. OpenAI has released a Secure MCP Tunnel — an outbound-only HTTPS solution that allows organizations to connect their private, on-premises Model Context Protocol (MCP) servers to OpenAI products without ever opening an inbound firewall port. If your organization has been watching MCP adoption from the sidelines because your internal data lives behind a firewall that your security team would never open to the outside world — that objection is now gone. ...

The moment fintech practitioners have been watching for has arrived: Robinhood has opened its platform to AI agents. With the launch of Agentic Trading and the Agentic Credit Card, Robinhood becomes the first major mainstream retail financial platform to give third-party AI agents direct, official access to trading accounts and spending power — no unofficial APIs, no workarounds required. This is a significant milestone for the agentic AI ecosystem. Moving money is one of the highest-stakes real-world actions an agent can take, and Robinhood has bet that safety controls and transparency can make it viable at consumer scale. ...

Simon Willison — one of the most credible and consistent independent voices in AI infrastructure and tooling — published a piece on May 27, 2026, making a claim that would have sounded optimistic 18 months ago but now feels like describing observable reality: Anthropic and OpenAI have both definitively found product-market fit. The post is worth reading carefully. Willison isn’t making a bullish prediction or a promotional argument. He’s doing what he does best: reasoning from data points he can actually observe, drawing careful inferences, and naming what’s happening. ...

Anthropic Expands Public Access to Claude Mythos AI Model — U.S. and Allied Government Partners Anthropic’s most capable and most restricted AI model is about to have a wider audience. The company announced on May 26, 2026 that it will expand access to Claude Mythos — its frontier AI model, which has been positioned as too powerful for general public release — to “critical partners including the U.S. and allied governments.” ...

Anthropic Opens Mythos in Claude Code as Agent Debate Sharpens Two things happened this week that, taken together, capture the moment AI and software development finds itself in: Anthropic opened up its most powerful and most restricted model inside Claude Code, and prominent voices in software engineering are openly debating whether autonomous AI agents represent a force multiplier for developers or an existential threat to how software actually gets built. ...

The standard story about feature flags is simple: hide a new feature behind a flag, gradually roll it out, watch metrics, rollback if needed. It’s a human-paced workflow designed around human review cycles. But AI agents writing and shipping code are operating at a fundamentally different tempo — and the tooling needs to catch up. Cloudflare’s Flagship, launched during Agents Week 2026 and continuing to gain traction, is a purpose-built answer to this problem. ...

The Model Context Protocol continues its march into every corner of the software ecosystem, and on May 26, 2026, it arrived in crypto. Coinbase’s Base Layer-2 network launched Base MCP — a native MCP integration that gives AI agents in Claude, ChatGPT, Cursor, and Gemini the ability to directly control crypto wallets and interact with DeFi applications. No custom integrations, no complex SDKs. Just plug in the MCP server and your AI agent can now send funds, swap tokens, check balances, and sign messages on Base L2. ...