Anthropic Silently Fixed a Second Claude Code Sandbox Bypass — 130 Versions, No CVE, No Advisory If you’ve been running Claude Code between version 2.0.24 and 2.1.89, your network sandbox wasn’t doing what you thought it was. For approximately 5.5 months — across roughly 130 versions — a null-byte hostname injection vulnerability allowed an attacker to bypass the network sandbox’s allowlist entirely, enabling arbitrary outbound connections and potential exfiltration of credentials, source code, or any sensitive data accessible from your sandbox environment. ...

How to Scale OpenClaw Agents on Kubernetes with ACP and acpx Running one OpenClaw agent on your laptop is easy. Running a hundred of them reliably, in parallel, across a production cluster — that’s a different problem entirely. At an AI Engineer event, Onur Solmaz, OpenClaw’s core maintainer at Hugging Face, showed exactly how to do it. The talk, titled “Scaling Agents on Kubernetes with acpx and ACP,” is the most authoritative take on production-grade OpenClaw infrastructure available right now — and it introduces a Go-based Kubernetes operator that makes horizontal agent scaling significantly more tractable. ...

How to Set Up Cloudflare CASB with Claude Compliance API for Enterprise Agent Governance Enterprise security teams finally have a native, agentless way to monitor what Claude is doing across their organization — and it doesn’t require touching endpoints, installing agents, or routing all traffic through a proxy. On May 21, 2026, Cloudflare announced that its Cloud Access Security Broker (CASB) now supports Anthropic’s Claude Compliance API. The integration gives security and compliance teams out-of-band visibility into Claude Enterprise and Claude Platform activity, directly in the Cloudflare dashboard. ...

JetBrains Koog Hits 1.0: Building Your First Production AI Agent in Kotlin If you’ve ever tried to build a reliable AI agent on the JVM and ended up with a spaghetti mess of LLM calls, retry logic, and unexplainable failures — JetBrains just dropped your solution. At KotlinConf'26 in Munich, Koog reached stable 1.0, and it came with a real-world production story: Mercedes-Benz is using it to power vehicle maintenance scheduling agents at dealers worldwide. ...

Zscaler Acquires Symmetry Systems to Lock Down AI Agent Identity — First M&A Deal Targeting Agentic Security The security industry just sent its clearest signal yet that AI agents aren’t a future concern — they’re a present one. On May 21, 2026, Zscaler announced its intent to acquire Symmetry Systems, a company that built exactly the kind of identity graph technology needed to see and control what AI agents are doing inside enterprise environments. ...

Eight days ago, Nebula Security’s AI agent Vega made headlines by discovering nginx-rift (CVE-2026-42945) — an unauthenticated remote code execution vulnerability in nginx 1.31.0 that sent administrators scrambling to patch. The nginx team released a fix. Teams updated. Incident closed. Except Vega kept digging. On May 21, Nebula Security disclosed nginx-poolslip — a second, distinct RCE vulnerability in nginx 1.31.0 that specifically bypasses the nginx-rift patch. There is currently no CVE assigned and no available fix. ...

When SpaceX quietly filed its S-1 IPO registration with the SEC this week, it handed the world a detailed financial X-ray of one of the most consequential private deals in AI infrastructure history. Buried inside: Anthropic agreed to pay SpaceX $1.25 billion per month for access to its Colossus data centers in Memphis, Tennessee — a commitment running through May 2029, totaling nearly $45 billion over three years. That number is staggering. For context: SpaceX reported $18.7 billion in total revenue for all of 2025. Anthropic’s monthly payment to Elon Musk’s company alone approaches what the rocket company earned in an entire quarter last year. ...

Design tooling had a genuinely remarkable 24 hours on May 20, 2026. Figma launched its native AI Design Agent into open beta. Google, at I/O 2026, unveiled Stitch — a streaming AI design agent with real-time multiplayer editing. Both tools dropped on the same day. And Google Stitch is free, while Figma’s full rollout will cost $15 per seat for Professional and above plans. The design agent wars are officially on. ...

One of the persistent gaps in enterprise AI agent development has been the lack of systematic safety tooling that fits into existing engineering workflows. Security reviews, red-team exercises, and deployment risk assessments are often manual, ad hoc, or entirely absent — because the tooling to make them repeatable simply didn’t exist. Microsoft is trying to fix that. The company announced two new open-source tools on May 20: RAMPART and Clarity — both MIT-licensed and available on GitHub today. ...

There’s a new heavyweight in the agentic AI harness space — and it just closed a round that industry watchers didn’t see coming. NanoClaw, the containerized AI agent harness built as a security-first alternative to OpenClaw, has raised a $12 million oversubscribed seed round at a $62 million post-money valuation. The round closed in just four days. And that’s the second surprise — the first was that founders Gavriel and Lazer Cohen had already turned down a ~$20 million acquisition offer before this round even closed. ...