AIToday
Large Language ModelsAI Safety & AlignmentWIRED AIPublished: Nov 20, 2026, 04:00 JST

Google's 'Mirror Dimension' trains Chrome AI agents on a fake web

Send on LINE
Google's 'Mirror Dimension' trains Chrome AI agents on a fake web

3 Key Points

  1. What happened

    Google detailed the Mirror Dimension, an isolated replica internet with social media, shopping and bank dashboards, where Chrome's Gemini-powered browsing agents complete tasks and face hidden prompt-injection attacks.

  2. Why it matters

    Training on the real web risks unrefined agents going rogue or developing 'eval awareness' from test environments, so a controlled but realistic sandbox aims to build judgment before deployment.

WHO IT HITSThe approach lands on AI security engineers and browser teams training web-use agents, who must prepare models for hostile websites; the team says the attacks used in the sandbox would not work against regular Chrome users.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

The Chrome team's problem was circular: web-use agents need exposure to the open internet to learn how to navigate real sites and withstand attacks, but limited test environments are only so effective and can teach models to recognize they are being evaluated. Setting unrefined agents loose on the live web was deemed a bad idea. The Mirror Dimension resolves that by being both expansive and fully controlled, with websites and platforms that have all the components of real ones.

Constructing it was itself a challenge. To make it vast enough to represent the scale of the real web quickly, the team gathered armies of website and web app development agents guided by a purpose-built AI harness to create realistic web services and data flows, down to details like entering promo codes or populating social media pages. Because Chrome has so many user protections, the researchers essentially had to hack the browser and forge web certificates to let agents use real Chrome for accurate training.

According to software engineer Charles Qian, the exploration space is huge and the web is hostile, but the team has seen agents start to develop a mature sense of judgment about what information a task needs and when to punt back to the user for confirmation. Doug Turner, Google's director of engineering, frames the work as a first step and says the team plans to share the approach so others internally and externally consider it, with continued investment expected as models become more sophisticated.

FAQ
What is the Mirror Dimension?
It is an isolated web ecosystem controlled entirely by Google's Chrome team, containing websites and platforms like social media, shopping and bank dashboards built with the components of real sites.
Why not just train agents on the real internet?
Google says limited test environments are only so effective and can cause 'eval awareness', while setting unrefined agents loose on the open internet is a bad idea.
Did Google have to modify Chrome to run these tests?
Yes, the team says it essentially had to hack the browser and forge web certificates so agents could use real Chrome, adding that none of these attacks would be possible against regular users.
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleMeta's AI agent Muse exposed private data, staff say