A groundbreaking technology has been developed by Anthropic, allowing its AI agent, Claude, to control computers and perform tasks independently. This feature, currently in "research preview" mode, gives Claude the power to navigate computer screens and interact with applications, files, and settings to run assigned tasks. The AI can open files, launch apps, click icons, type text, and browse the web on its own, making it a significant evolution of AI agents.
The new skill is limited to paid plans, specifically Claude Pro and Max subscribers, and is only accessible on Mac through the Claude Mac app. Free users and those with Team or Enterprise plans are not eligible for this feature. Additionally, Claude will try to avoid interacting with certain apps or sensitive data, such as banking, financial, healthcare, or legal matters.
However, users must still exercise caution when granting Claude permission to access certain apps or files. The AI can capture screenshots to learn how to navigate the computer, which means it can see and record whatever information is on the screen. Users should ensure that any apps or files with confidential or private data are not open and visible when issuing a command.
This technology has the potential to save time and labor, but it also poses risks, especially given the unpredictability of AI bots. To mitigate these risks, Anthropic advises users to start with simple tasks and fine-tune their instructions to avoid any type of misunderstanding.
The development of AI agents that can control computers independently marks a significant milestone in the evolution of artificial intelligence. While this technology has the potential to revolutionize the way we work, it also raises important questions about security and oversight. As Nigerian tech professionals and developers continue to explore the possibilities of AI, it is essential that they prioritize caution and carefully evaluate the potential risks and benefits of this technology.
