Artificial intelligence company, Anthropic, has introduced a major upgrade to its Claude chatbot, enabling it to use computers and complete tasks on behalf of users.

The new capability allows Claude to move beyond answering prompts and actively interact with software interfaces, marking a significant step in the evolution of so-called AI agents.

With the update, Claude can navigate a computer much like a human user. It moves a cursor, clicks through menus, fills out forms, and completes multi-step workflows across applications.

This update transforms Claude from a passive assistant into an active operator that executes instructions end-to-end, rather than simply suggesting what a user should do next.

Read also

The feature builds on Anthropic’s broader push into “agentic AI”, where developers design systems to plan, adapt, and act independently within defined tasks.

The company says the new functionality improves real-world productivity by allowing users to delegate routine digital work.

Instead of manually switching between tools, users can instruct Claude to complete tasks such as data entry, research, or navigating workflows across different applications.

ALSO WATCH MARKETING EDGE ONTV

Industry analysts see this as part of a wider shift across the AI sector, where companies are racing to develop systems that perform tasks, not just generate text.

Anthropic’s move comes amid intensifying competition in the AI space, with rivals including OpenAI and Google investing heavily in autonomous AI systems.

These tools increasingly handle complex workflows, from coding to business operations, raising expectations for what AI assistants can deliver.

However, experts have also raised concerns about reliability and safety, warning that systems capable of acting autonomously could introduce new risks if not properly controlled.

Anthropic is currently releasing the computer-use feature in a limited preview, with access restricted to select users and developers.

The company says it will continue refining the system based on feedback, particularly around accuracy, safety, and task execution reliability. This update signals a broader shift in how companies deploy AI.