background agent
In plain wordsAn AI agent you hand a task to and then leave alone: it works by itself in the cloud, often for many minutes or hours, and reports back with finished work for you to review.
An AI agentAI agent, n.: An AI system that does not just answer but acts: given a goal, it decides on steps, uses tools such as a browser or code editor, checks the results and keeps going until the job is done., most often a coding agentcoding agent, n.: An AI agent that writes and fixes software by itself: it reads a codebase, edits files, runs commands and tests, and keeps working until a feature is built or a bug is fixed., that runs asynchronously and without live supervision, typically in a cloud sandboxsandbox, n.: A sealed-off computer space where an AI agent can run programs and change files without being able to harm the real system, reach private data or contact the outside world unless allowed. or virtual machine provisioned with a copy of the user's repository, while the user does other work. A task is assigned from an issue tracker, chat, mobile app or editor; the agent sets up the environment, makes changes, runs tests and returns its results as a branch or pull request with logs for review. Running many such agents in parallel is a defining use. Because nobody approves individual actions, background agents rely on environment-level controls — restricted network access, scoped credentials, setup scripts and reviewable outputs — rather than per-command permission promptspermission system, n.: The rules that decide what an AI agent may do on its own, what it must ask a person about first, and what it may never do — like a phone's settings for which apps can use the camera..
I assign the flaky-test tickets to background agents on Friday and review their pull requests on Monday.
Origin
Cognition's DevinDevin, prop. n.: A commercial AI 'software engineer' from the startup Cognition, unveiled in 2024, that takes on programming tasks in its own cloud workspace, complete with a web browser, a code editor and a command line. (March 2024) already worked asynchronously in its own cloud environment. Cursor named a preview feature 'Background Agent' in version 0.50 (15 May 2025); OpenAI's CodexCodex, prop. n.: A name OpenAI has used twice: first for a 2021 AI model that wrote code and powered the original GitHub Copilot, then from 2025 for its family of coding agents. cloud agent (16 May 2025), GitHub's Copilot coding agent (19 May 2025) and Google's Jules (public beta, May 2025) launched within days of one another, establishing the category.
Usage note
Neighboring terms include 'cloud agent' and 'ambient agent' (one triggered by events rather than by chat); product names change often. Background agents trade immediacy for parallelism: mistakes surface at review time rather than as they happen.
Related
- coding agentcoding agent, n.: An AI agent that writes and fixes software by itself: it reads a codebase, edits files, runs commands and tests, and keeps working until a feature is built or a bug is fixed.
- agentic codingagentic coding, n.: Writing software by directing AI agents that plan, edit and test code themselves, with the human acting more as a reviewer and manager than as the person typing every line.
- sandboxsandbox, n.: A sealed-off computer space where an AI agent can run programs and change files without being able to harm the real system, reach private data or contact the outside world unless allowed.
- long-horizon tasklong-horizon task, n.: A job that takes many steps and a long time — hours or days of work rather than a single answer — so an AI must stay on track and recover from its own mistakes.
- CodexCodex, prop. n.: A name OpenAI has used twice: first for a 2021 AI model that wrote code and powered the original GitHub Copilot, then from 2025 for its family of coding agents.
- DevinDevin, prop. n.: A commercial AI 'software engineer' from the startup Cognition, unveiled in 2024, that takes on programming tasks in its own cloud workspace, complete with a web browser, a code editor and a command line.
- copilotcopilot, n.: An AI assistant built into software that works alongside a person — suggesting code, drafting text or answering questions — while the person stays in charge, like a co-pilot beside the captain.
Sources
- Cursor (2025). Changelog 0.50: Background Agent (preview). 15 May 2025. cursor.com/changelog/0-50
- GitHub (2025). GitHub Copilot coding agent in public preview. GitHub Changelog, 19 May 2025. github.blog/changelog/2025-05-19-github-copilot-coding-agent-in-public-preview
- OpenAI (2025). Introducing Codex. 16 May 2025. openai.com/index/introducing-codex
Mentioned in
coding agentcoding agent, n.: An AI agent that writes and fixes software by itself: it reads a codebase, edits files, runs commands and tests, and keeps working until a feature is built or a bug is fixed., agentic codingagentic coding, n.: Writing software by directing AI agents that plan, edit and test code themselves, with the human acting more as a reviewer and manager than as the person typing every line., Claude CodeClaude Code, prop. n.: A coding assistant from Anthropic that works as an agent: given a programming task, it explores the project, changes the code and runs programs to check its work, from the web, a desktop app, an editor or a terminal., CodexCodex, prop. n.: A name OpenAI has used twice: first for a 2021 AI model that wrote code and powered the original GitHub Copilot, then from 2025 for its family of coding agents., long-horizon tasklong-horizon task, n.: A job that takes many steps and a long time — hours or days of work rather than a single answer — so an AI must stay on track and recover from its own mistakes., sandboxsandbox, n.: A sealed-off computer space where an AI agent can run programs and change files without being able to harm the real system, reach private data or contact the outside world unless allowed., agent swarmagent swarm, n.: A large group of AI agents working at the same time on parts of a problem, usually with little central control — a name loosely borrowed from swarms of insects., clawclaw, n.: Nickname for an always-on personal AI assistant, like OpenClaw, that lives on your own computer and does jobs for you whenever you message it., DevinDevin, prop. n.: A commercial AI 'software engineer' from the startup Cognition, unveiled in 2024, that takes on programming tasks in its own cloud workspace, complete with a web browser, a code editor and a command line., sleep-time computesleep-time compute, n.: Letting an AI think about your documents or project while it would otherwise sit idle, before you ask anything, so it can answer later questions faster and more cheaply.