OpenClaw update simplifies setup and browsing

OpenClaw, an open-source AI agent, has released version 2.0, its largest update since launching last November. The new version, also known as 2026.8.1, was shipped over the weekend and boasts 933 contributors, with 569 of them being first-timers, and over 16,000 merged pull requests.

Community manager Hannes Rudolph explains that the team initially set out to simplify installation and rebuild the browser app, but ended up reworking the entire codebase and releasing a full update. This update is a significant milestone for the project.

The new version aims to improve the user experience with a guided install that scans the user’s machine for existing AI access, such as Codex, ChatGPT, or Claude logins, and confirms the chosen model can answer a request before saving it.

Most configuration has been pushed out of the first run, allowing users to finish setting up the agent by talking to it. The browser Control UI is now chat-first, with the old overview page removed and sessions moved to a sidebar.

Files, approvals, terminals, git diffs, and a dockable browser panel sit alongside the conversation. The biggest addition to OpenClaw 2.0 is shared cloud sessions, which allow an admin to grant read-only, suggest, draft, or full-participation access to a live agent session with context.

Security has also been improved, with credential requests masked, an optional 1Password broker, and plugin provenance shown before install. One-time approvals that expire when an automation’s behavior changes have also been added, along with workspace-restricted filesystem access.

These improvements aim to make OpenClaw a more secure and user-friendly option for those looking for an open-source AI agent. OpenClaw’s user base has shifted towards a smaller, more committed group, with many users utilizing the tool for specific tasks such as inbox monitoring and scheduled network scans.

They face competition from other AI agents, such as ChatGPT Work and Claude Cowork. However, OpenClaw still offers ownership and control, allowing users to run the agent on their own hardware, with their own model, and their own data, without a subscription.

This aspect of OpenClaw is something that the competition can’t copy, and it remains a key differentiator for the project. As the AI environment continues to evolve, OpenClaw 2.0 aims to provide a more user-friendly and secure experience for its users.

In the middle of this update, it’s clear that OpenClaw is no longer the phenomenon it once was, but rather a reliable tool for a dedicated group of users. The project’s move to a nonprofit foundation in July and the publication of guides by AMD and Nvidia for running it locally on their hardware demonstrate its growing infrastructure and community support.

For those interested in learning more about OpenClaw and its capabilities, the project’s open-source nature and community-driven development make it an interesting option to explore.

The release of OpenClaw 2.0 marks a significant milestone for the project.

Its improved setup and browser app, along with shared cloud sessions and security improvements, make it a more attractive option for users looking for an open-source AI agent.