Google’s Agent Smith Is Quietly Changing Coding
Google’s internal Agent Smith automates coding tasks, works in the background, and is already so popular access had to be limited.

Google’s internal AI tool Google calls Agent Smith is already busy enough that access had to be restricted. The tool, which launched earlier this year, can automate coding work, run asynchronously, and let employees check in from their phones while it keeps working in the background.
That matters because Smith is not a demo or a side project. It builds on Google’s internal agentic coding platform, Antigravity, and it has enough access to internal tools and employee profiles to do more than autocomplete a few lines of code. For a company pushing hard on AI adoption, this is a glimpse of what day-to-day engineering work may look like inside Google next.

What Agent Smith actually does
Get the latest AI news in your inbox
Weekly picks of model releases, tools, and deep dives — no spam, unsubscribe anytime.
No spam. Unsubscribe at any time.
Based on the reporting, Smith is an inter

That asynchronous behavior is the part worth paying attention to. Most coding assistants still feel like enhanced autocomplete tools. Smith sounds closer to a delegated coworker that can continue a task after you close your laptop.
Google has not published a product page for Smith, but the tool appears to fit squarely into the company’s broader push around agentic AI. For developers, that usually means software that can plan, act, and update its own state across several steps rather than answering one prompt at a time.
- Smith launched earlier in 2026, according to people familiar with it.
- Access was restricted after internal demand surged.
- It can run in the background without an active laptop session.
- Employees can send instructions from their phones.
- It can pull documents tied to employee profiles.
The big practical difference is time. A tool that can keep moving while you are in meetings or away from your desk changes the shape of a workday. Instead of waiting to resume a task, an engineer can return to a partially completed workflow and decide what happens next.
Why Google is pushing agents so hard
Google’s internal push did not appear out of nowhere. CEO
// Related Articles
- [AGENT]
How to Switch AI Outputs from Markdown to HTML
- [AGENT]
Anthropic’s Cat Wu on proactive AI assistants
- [AGENT]
How to Run Hermes Agent on Discord
- [AGENT]
Why RAGFlow is the right open-source RAG engine to self-host
- [AGENT]
How to Add Temporal RAG in Production
- [AGENT]
GitHub Agentic Workflows puts AI agents in Actions