Computer use
- Definition
- An AI capability where a model can directly interact with a computer's graphical interface, clicking buttons, typing text, navigating menus, and reading screen content just like a human user would.
- Why it matters
- Computer use is the bridge between AI capability and real-world action. Most enterprise software was never designed for API integration: it has GUIs, web portals, and legacy desktop applications. An AI that can operate these interfaces like a human bypasses the entire integration bottleneck. This is transformative for automation: instead of building custom API connectors for every system, you point an agent at the screen and let it work. The implications for RPA (robotic process automation), a $3B+ market, are existential. AI-powered computer use is more flexible, more intelligent, and cheaper than traditional RPA bots, and it can handle edge cases that scripted automation cannot.
- In practice
- Anthropic launched computer use as a beta feature for Claude in October 2024, enabling the model to control a desktop environment through screenshots and mouse/keyboard actions. OpenAI followed with Operator, a browser-based agent that completes tasks like filling out forms and placing orders. Google's Project Mariner similarly controls Chrome to accomplish web tasks. In enterprise settings, companies are using computer use to automate workflows in legacy systems like SAP, Oracle, and custom internal tools that lack modern APIs, reporting 60-80% time savings on repetitive GUI-based tasks.
We cover products & deployment every week.
Get the 5 AI stories that matter — free, every Friday.
Related terms
Agent
An AI system that can autonomously plan, use tools, and execute multi-step tasks on behalf of a user. Agents are the next major product paradigm after chatbots, with every major lab shipping agent frameworks.
Tool use
The ability of an AI model to invoke external tools, such as web search, code execution, or database queries, to augment its capabilities. Tool use transforms models from knowledge stores into action-taking agents.
Agentic workflow
A multi-step process where an AI agent plans, executes, evaluates, and iterates on tasks with minimal human intervention. Unlike single-turn prompts, agentic workflows involve loops, branching logic, and tool calls that unfold over minutes or hours.
Co-pilot
An AI assistant that works alongside a human user within an existing workflow, providing suggestions, automating sub-tasks, and augmenting productivity while keeping the human in control of final decisions.
Know the terms. Know the moves.
Get the 5 AI stories that matter every Friday — free.
Free forever. No spam.