# Why give an AI agent its own computer?

An AI agent with its own computer works apart from your devices, keeps going when you close your laptop, keeps its files and sign-ins between tasks, and can be rolled back when it makes a mistake.

Updated 2026-11-18. Page: https://casama.ai/learn/why-give-an-ai-agent-its-own-computer

## Three places an agent can work

| | Your own computer | A browser session | A computer of its own |
|---|---|---|---|
| Setup | Install an app | None | None, or self-host |
| Can reach your files | Yes | No | Only what you share |
| Works while you're away | Only if your machine stays on | For a short session | Yes |
| Keeps files and sign-ins | Yes | Usually not | Yes |
| If it makes a mistake | It happened on your machine | Limited to the session | Limited to its machine; restore a checkpoint |
| Can run code and install tools | Yes | No | Yes |

## What an own computer makes possible

- **Long work.** A task that takes an hour doesn't need your laptop open for an hour.
- **Real tools.** A full machine has a shell, a file system and a browser, so the agent can
  download a file, convert it, run a script on it and upload the result.
- **A working state.** The half-finished spreadsheet, the cloned repository and the signed-in
  session are still there next time.
- **Containment.** The agent can't read your documents or your saved passwords, because they
  aren't on its machine.
- **Undo.** A snapshot before a risky command lets you put everything back.
- **Watching.** You can open the computer's screen, see what the agent is doing, and take over.

## The costs

- **Money.** A computer that runs costs something per hour. Good systems put it to sleep when
  idle and wake it with everything intact.
- **Sign-ins.** The agent's computer isn't signed in to your accounts. You connect apps, or hand
  it logins through a password manager, ideally in a way the AI model never sees.
- **Bot checks.** Some websites challenge cloud browsers. Being able to take over for a moment
  helps.

## One computer for all agents, or one each?

Some products give every agent its own machine. Others run all of a person's agents on one shared
machine. One each means a research agent that installs odd software can't break the assistant that
handles your mail, and each can be limited to the sites it needs.

*Diagram: Two diagrams side by side. On the left, three bots all feed into one shared computer that holds one set of files and a shared setup, so what one bot does can affect the others. On the right, three bots each have their own computer with their own files and sign-ins, so each bot is walled off from the others, as in Casama.*

Bots on one shared computer share its files and setup. In Casama, each bot works on its own computer.

## How Casama does it

Every Casama bot works on its own isolated Linux machine with a shell, files and a browser, and
optionally a full desktop. Computers sleep when idle and wake with everything intact. You can watch
one live, take control, and restore a checkpoint taken before a destructive command. Logins from
1Password or Bitwarden are filled in on the way out and never shown to the model. If you'd rather a
bot used your own machine for a task, you can lend it through the desktop app, where high-risk
actions always ask.

## Related questions

- [How do AI agents work?](https://casama.ai/learn/how-do-ai-agents-work)
- [What is a computer use agent, and which let you take over?](https://casama.ai/learn/what-is-a-computer-use-agent)
- [Is it safe to give an AI agent access to my email?](https://casama.ai/learn/is-it-safe-to-give-an-ai-agent-access-to-my-email)

Join the waitlist at https://casama.ai/waitlist (Casama Cloud is invite-only for now), or run it on your own server: https://docs.casama.ai/self-hosting/
