OpenClaw shipped v2026.3.22 today, and it’s the biggest release the platform has seen in months. If you’ve been waiting for a native plugin ecosystem, longer-running agents, or GPT-5.4 access, this is the update you’ve been waiting for. Here’s what’s new — and what breaks. The Headline Feature: ClawHub Plugin Marketplace OpenClaw now has a native plugin marketplace called ClawHub. When you run openclaw plugins install <package>, it checks ClawHub first, falling back to npm only when ClawHub doesn’t have that package or version. ...

Anthropic just crossed a threshold that a lot of AI observers have been waiting for: its Claude AI can now directly control your computer. The company announced today that Claude Cowork — along with Claude Code — is being updated to perform tasks using your Mac or PC, opening files, running browser sessions, and executing multi-step workflows without you having to hold its hand at every step. This is no small shift. Computer-use AI has been a proof-of-concept for a while, but Anthropic is now putting it into the hands of paying subscribers. If you’re on Claude Pro or Claude Max, you can activate this on macOS today. ...

OpenClaw v2026.3.22 ships ClawHub — a native plugin and skills marketplace built directly into the platform. No more hunting npm packages or manually dropping files into skill directories. ClawHub lets you browse, install, and manage extensions from a curated registry in one place. This guide walks you through everything: how ClawHub works, how to install your first skill, and a real-world example using Hitem3D — the first major third-party integration spotlighted on the marketplace — to generate 3D models from images. ...

Meta just made its most pointed bet yet on personal AI agents — and it came with a familiar face. The company has acqui-hired the full team behind Dreamer, a startup building tools for creating personalized AI agents, in a deal that brings former Meta and Google executive Hugo Barra back into the Meta fold. The Dreamer team will join Meta Superintelligence Labs, working under Chief AI Officer Alexandr Wang — formerly the founder of Scale AI — according to a Bloomberg report confirmed by multiple sources and the founders themselves on X. ...

OpenClaw shipped version 2026.3.22 today, and the release notes read like the team decided to clear out a year’s worth of backlog in one shot. There’s a native plugin marketplace, a long-overdue timeout fix, expanded model support, a substantial security patch batch, and the final closure of the MoltBot rebrand saga. This one has breaking changes — read carefully before you update. The Headliner: ClawHub Plugin Marketplace The biggest architectural addition in this release is ClawHub — a native plugin marketplace baked directly into the OpenClaw runtime. Previously, extending OpenClaw meant hunting down npm packages or manually managing skill directories. ClawHub centralizes this: you can now browse, install, and update skills and plugins from a curated registry without leaving the tool. ...

OpenClaw exploded onto the scene in November 2025 and became, by any measure, the fastest-growing open-source project in history. Within months, tens of millions of people were using it to automate their lives — running shell commands, managing files, connecting to messaging platforms, building new agent skills overnight. NVIDIA CEO Jensen Huang called it “the operating system for personal AI.” But explosive growth brings explosive risk. And on March 23, 2026, at RSA Conference in San Francisco, Cisco decided to do something about it. ...

For years, the promise of enterprise AI was efficiency at scale — machines doing the work of many. What the Wall Street Journal reported this weekend is something more interesting: machines doing the work of management itself. Mark Zuckerberg is building a personal AI agent. And he’s already using it. What the WSJ Report Says According to sources cited by the Wall Street Journal on March 22, Zuckerberg’s agent is designed to do something simple but structurally significant: get him information faster by cutting out the layers of people who would normally surface it. ...

RSA Conference opened its doors in San Francisco on March 23, 2026, and the message was impossible to miss before the first keynote even ended: agentic AI security is now the primary concern of the entire enterprise security industry. For years, RSAC has been the place where the security industry takes collective stock of where threats are heading. This year, every major vendor arrived with the same answer: AI agents are the new attack surface, and the industry is barely ready for it. ...

There’s a wall every engineering team hits when they try to move AI agents from demo to production: identity and access management. An agent needs credentials to do anything useful — database access, API keys, infrastructure permissions. But credentials that live inside an agent are credentials that can be leaked, stolen, or misused. Traditional IAM wasn’t designed for ephemeral, autonomous software actors. And so most production agent deployments end up making one of two bad choices: over-permissioned agents with broad access they don’t need, or under-permissioned agents so locked down they can’t do their jobs. ...

OpenClaw went viral for a reason — it’s the closest thing to a real personal AI operating system most developers have ever touched. But as Cisco’s own engineers put it at RSA Conference 2026 this week: the fastest-growing open source project in history is also a massive target. Their answer is DefenseClaw, an open-source security framework built specifically for OpenClaw deployments. What DefenseClaw Actually Does Cisco unveiled DefenseClaw on Monday at RSAC 2026, the San Francisco security conference that this year has turned almost entirely toward AI agent security. The framework ships with six distinct components designed to close the security gap that’s opened up as OpenClaw adoption has exploded: ...