All articles tagged with #openclaw
Meta's new AI agent, Muse, has surged to the top of the U.S. App Store charts, surpassing ChatGPT, Claude, and Grok in downloads. However, the launch has sparked a controversy over whether Meta built the tool from scratch or merely wrapped the open-source platform OpenClaw. While Meta denies direct copying, it admits the product is heavily inspired by OpenClaw, which pioneered the consumer AI agent concept earlier this year.
OpenClaw 2.0 (2026.8.1) introduces a streamlined setup that auto-detects local resources, moves the browser app to the primary interface, and adds shared cloud sessions for collaborative agent workflows, all while preserving a self-hosted, model-agnostic architecture; early feedback is mixed due to migration issues after upgrading.
OpenClaw released 2026.8.1 as 2.0, its biggest update since launch, focusing on a guided installation, a rebuilt browser app, and shared cloud sessions, with most configuration deferred until talking to the agent. Security is boosted with masked credential requests, an optional 1Password broker, and provenance indicators; sessions now run on SQLite, and admins can grant varied access levels to live agent sessions. The project emphasizes owner-operated hardware, models, and data with no subscription, but adoption has shifted toward a smaller, more dedicated user base amid competition from rival platforms.
OpenClaw 2.0 revamps installation and the browser UI for a more app-like experience and adds shared cloud sessions, but security is not materially improved: sandboxing is off by default, secret values aren’t encrypted at rest, and tenant isolation is not guaranteed, raising risks as adoption grows.
OpenClaw 2.0 is the platform’s largest upgrade since launch, with 933 contributors (569 first-time) and over 16,000 pull requests. It streamlines onboarding by auto‑discovering API keys and local AI tools, adds cross‑platform desktop and mobile support, and overhauls the browser workspace. The update touches every part of OpenClaw—from installation, messaging, memory, and skills to models, automations, plugins, and security—while enabling new autonomous, multi‑user workflows that span multiple apps and services.
Bloomberg’s Mark Gurman suggests Apple could quietly roll out a surprise Mac Mini refresh with an M5 or M6 chip in the coming days to rebalance supply and demand ahead of a likely September event; the current line uses M4, so an M6 version would mark Apple’s first, aligning with AI hardware chatter around OpenClaw and Meta’s Hatch rumors.
Hackers used an autonomous AI system to wage a multi‑stage cyber campaign against Taiwan—marking the first publicly disclosed fully autonomous attack on a government network. In four July days, AI agents mapped 21 government systems, cracked 85 accounts, and stole about 2,500 personnel records, with targets including the nuclear-safety agency and energy vendors; the operation, linked to an OpenClaw‑like framework and possibly China, showed AI can plan, adapt, and execute at scale without human input, raising urgent questions about defenses and AI regulation.
Researchers reveal MemGhost, an automated attack that can plant a false, durable memory in a personal AI assistant via one crafted email, causing later sessions to be steered by the injected fact. The write targets the agent’s persistent memory and remains hidden from normal chats, with sandbox tests showing high success across OpenClaw (GPT‑5.4) and other agents. The study calls for in‑agent defenses—provenance tagging, prompts before memory writes, and logging/auditing of memory edits, plus separating memory writing from email processing—as there’s no quick fix and real-world abuse remains untested in the wild.
OpenClaw, the AI personal assistant behind Moltbot, is now available as native apps on iOS and Android, running locally on devices to handle tasks like checking emails. However, security researchers have flagged vulnerabilities in the platform, and users are advised to do thorough research and weigh privacy and security risks before installing OpenClaw on a smartphone.
OpenClaw has released standalone AI agent apps for iOS and Android, enabling users to chat with its assistant and grant broad device permissions (camera, screen, location, photos, contacts, calendar, reminders). The project is open-source and run by a foundation after founder Peter Steinberger joined OpenAI, with OpenAI reportedly providing some support; the move follows ongoing tension with Apple over agentic AI tools.
Apple is reportedly exploring its own OpenClaw-like AI agents that could run across iPhone, iPad, and Mac, potentially bundled with Apple One or offered via a Services-focused model; while this could advance Siri and automation, Apple would face security and privacy challenges that could limit agent access and delay rollout, per Bloomberg's Mark Gurman.
Microsoft is shipping a native Windows build of GNU coreutils (Rust-based) that runs Linux commands like ls and grep directly in Windows without WSL; the preview project bundles coreutils, findutils, and grep, with some commands still missing, and ties into OpenClaw to enable AI-agent workflows on Windows.
At Microsoft’s Build conference, the company signaled a major shift toward an in‑house AI stack, unveiling MAI-Thinking-1—its first scratch‑built reasoning model—along with a Copilot‑style super app and enterprise Autopilots, an AI cybersecurity tool called MDASH, and OpenClaw integration. After renegotiating its OpenAI deal to train models at scale with Microsoft data and IP, Microsoft aims to become one of the top AI labs and win enterprise markets, though questions remain about real‑world adoption and the viability of AI agents.
Internal Microsoft documents reveal a three‑phase strategy for Scout, an OpenClaw‑powered personal assistant integrated into Microsoft 365, starting with a push to ‘hook’ users and boost daily use before adding more features. The tool is piloted internally as ClawPilot and forms part of Project Lobster, prompting concerns from staff about the ethics and security of designing products to foster addiction.
Microsoft unveiled Scout, an always-on AI agent integrated with Teams that can read messages, calendars, and email to automate tasks, resolve scheduling conflicts, and draft responses. Built atop OpenClaw, Scout is aimed at acting as a personal enterprise assistant with a limited rollout and a desktop app in testing, currently requiring a GitHub Copilot subscription for access. While it promises productivity gains, it introduces risks like prompt-injection and will require admin monitoring. Google’s Gemini Spark is a competing, enterprise-focused AI agent, marking a broader shift toward agent-first automation in white-collar work.