This agent has the capabilities of AI research along with the new ability to take control of web browsers, computer files, and software such as spreadsheets and slide decks. Open AI has acknowledged that giving AI control over computer systems comes with increased risks, so they are taking preventative measures to ensure no agents enable the creation of biological hazards. The model is trained to confirm with the user before it takes any destructive or irreversible action, and users can interrupt, take over the browser, or stop tasks at any point.