Anthropic's Claude Can Now Control Your Computer Claude, the AI model developed by Anthropic, now has the ability to take control of a user’s computer to perform tasks such as retrieving files stored on a hard drive. The feature, announced by Anthropic, requires users to be on a qualifying subscription plan to access it. This development comes as part of Anthropic’s efforts to enhance Claude’s agentic capabilities, allowing the AI to act more independently compared to traditional chatbots. The release follows the rapid rise of the open-source OpenClaw framework, which has enabled AI tools to execute simple commands autonomously on computers or integrated systems. OpenClaw has spurred the creation of various "claws," or AI-driven tools, capable of performing tasks with minimal user input. Nvidia recently introduced NemoClaw, a framework designed to simplify the setup and installation of OpenClaw, though it includes security settings to mitigate risks. Anthropic’s new feature allows Claude to interact with a computer by leveraging connectors with applications like Google Calendar or Slack. If the necessary tools or connectors are unavailable, Claude can manually perform tasks by simulating keyboard and mouse actions. The AI can use programs such as web browsers, development tools, and file-opening applications to carry out these tasks. However, the system will always prompt the user for permission before executing any action, and users can stop the process at any time. While the ability to control a computer offers convenience for specific tasks, it also raises security concerns. Experts warn that agentic AI systems, like Claude, could potentially take significant or abrupt actions without prior warning, increasing the risk of unintended consequences.#nvidia #openclaw #anthropic #claude #dispatch
