Meta Admits Muse Was Heavily Inspired by OpenClaw in AI Copycat Controversy

TL;DR
- Meta has admitted its new Muse AI assistant was heavily inspired by open-source agentic tool OpenClaw after developers found identical workspace filenames, prompts, and config content.
- Meta claims Muse was built from scratch for its own infrastructure, but concedes its team studied and borrowed concepts and file structures from OpenClaw during rapid prototyping.
- The controversy has sparked a major debate over Big Tech borrowing from open-source AI, licensing ethics, and whether trust between indie builders and tech giants can survive.
How The Internet Caught Meta Copying
The controversy erupted within hours of Meta quietly rolling out Muse, its new autonomous AI assistant designed to run local workflows, control browsers, and manage files. Developers who downloaded the early build quickly noticed something strange: the workspace directory looked eerily familiar.
Filenames like SOUL.md, USER.md, MEMORY.md, and TOOLS.md — signature files used by OpenClaw, the popular open-source personal AI assistant created by Peter Steinberger — were present in Muse with nearly identical names and structures. Even default content, system prompts, and onboarding language appeared to be lifted almost verbatim.
Side-by-side screenshots posted on X, GitHub, and Hacker News went viral, with open-source maintainers pointing out matching markdown headers, identical memory-file logic, and similar agent loop instructions. For many, it was too close to be coincidence.
Meta's Admission: Inspired, Not Stolen
After 48 hours of mounting pressure, Meta responded. In a statement shared by a Meta engineering lead, the company acknowledged that Muse was heavily inspired by OpenClaw.
Meta admitted its prototype team had extensively studied OpenClaw as a leading example of open-source personal agents and that early internal builds referenced its workspace conventions. The company argued this was common practice in product research, and insisted the final version of Muse was built from scratch on Meta's own agent stack with proprietary models, sandboxing, and enterprise integrations.
In short: Yes, we looked closely and borrowed ideas and file patterns, but no, we didn't just fork the code.
OpenClaw Creator Responds
Peter Steinberger, the creator of OpenClaw, confirmed he had been contacted by Meta after the findings went public. While stopping short of accusing Meta of license violation — OpenClaw is permissively licensed — he expressed frustration that a trillion-dollar company would copy filenames and content without attribution or collaboration.
He noted that inspiration is welcome in open source, but direct copying of user-facing files without credit crosses an ethical line. Steinberger said he would welcome Meta contributing back upstream rather than simply absorbing community work.
Built From Scratch? Developers Aren't Convinced
Meta's built-from-scratch defense has done little to calm critics. Independent engineers who decompiled Muse point out that beyond filenames, behavioral patterns like how the agent bootstraps memory, handles daily notes, and structures tool permissions mirror OpenClaw's architecture almost exactly.
Legal experts say Meta is likely safe on copyright grounds since ideas and functional filenames are hard to protect, especially under permissive open-source licenses. But reputationally, the damage is growing, with critics accusing Meta of fast-following instead of innovating in the race against OpenAI, Anthropic, and Google.
What This Means For Open-Source AI Trust
This is bigger than a few markdown files. The Muse vs. OpenClaw saga has become a flashpoint for a deeper fear in the AI community: that Big Tech will treat open-source labs as free R&D.
For indie builders, the message is mixed. On one hand, having Meta validate your design is proof you're building the future. On the other, seeing your work shipped inside a closed competitor without credit or compensation feels like exploitation.
Expect renewed calls for stricter attribution requirements, defensive licensing like Commons Clause or Functional Source licenses, and louder demands for Big Tech to contribute code, funding, or partnerships back to the projects they learn from.
What's Next For Muse And OpenClaw
Meta says Muse will continue to evolve away from its OpenClaw-inspired origins, with upcoming releases set to overhaul the workspace format and introduce native Meta integrations for WhatsApp, Instagram, and Quest. The company has also hinted it may open-source parts of Muse's agent framework.
Meanwhile, OpenClaw has seen a massive surge in GitHub stars, contributors, and donations since the controversy — the classic Streisand effect. What started as a copycat scandal may ultimately cement OpenClaw as the standard for personal AI agents, forcing Meta to either collaborate or compete even harder to catch up.
Get All The Latest Updates Delivered Straight To Your Inbox For Free!