A Free Personal AI Agent Goes Viral, and Lawyers Start Asking Who Is Responsible for What It Does
OpenClaw, an open-source assistant that runs on your own computer and acts on your behalf, became one of the fastest-growing projects in software. It also raised a question nobody has answered: when an agent makes a decision, who is accountable?
An open-source project called OpenClaw has become, in a matter of weeks, one of the most talked-about pieces of software in the world. Built by an independent developer, it is a personal AI agent that runs on your own machine: you give it a goal and access to your tools, and it works — reading your files, browsing, sending messages, managing tasks — without a company in the middle. It is free, anyone can inspect how it works, and it passed 100,000 endorsements on the site where programmers share code, a level very few projects ever reach.
The appeal is obvious. Every major lab sells an assistant that lives on the lab's computers, sees what you tell it, and does what the lab permits. OpenClaw lives on yours.
The complications are the interesting part, and they are arriving faster than the answers. Legal commentators have spent the past weeks working through what happens when one of these agents acts in a way that matters — moves money, agrees to something, makes a commitment on your behalf. The agent has no legal personhood. It cannot be sued in its own name. It holds no title to anything. Its instruction to act in your interest is, as one analysis put it, a system prompt rather than an obligation a court can enforce. So the responsibility falls somewhere among the person who ran it, the person who wrote it, and nobody.
That is a live problem, not a philosophical one. Software that acts autonomously on an ordinary person's behalf is now free, popular, and running on home computers, and the framework for what happens when it goes wrong does not exist yet. The debate this has kicked off — about whether AI agents need some form of legal standing, and what proof would justify granting it — is the first time the personhood question has been forced by a practical mess rather than a thought experiment.