← Home

Chrome Now Uses Gemini to Autofill Forms with Data from Your Gmail

Chrome Crossed the Line: Gemini Now Reads Your Gmail to Autofill Forms

Chrome broke through. Not with a tsunami roar — but with a single click on "Suggested by Gemini."

After months of testing in Canary and promising the feature since May, Google activated Chrome Enhanced Autofill for all US users of Gemini with Personal Intelligence enabled. The pitch seems modest at first glance: when filling out a web form, Chrome can now automatically suggest information pulled directly from your Gmail inbox — order numbers, tracking codes, flight reservations, hotel details.

Behind this "UI detail" lies a deeper strategic shift. Chrome's autofill has always been a basic convenience tool, limited to static data the user had typed themselves: name, address, credit card number. Enhanced Autofill is the first time Chrome lets Gemini read your Gmail in search of relevant data. It marks the transition from passive autofill to active autofill, with AI acting as an intermediary between two Google products.

How it works — and what it reads

When you click a text field on any website, Chrome displays normal autofill suggestions (those saved previously) and, where applicable, a "Suggested by Gemini" section. The suggestions are generated locally by the Gemini model with Personal Intelligence, which scans Gmail for recognizable patterns — purchase confirmations, receipts, itineraries — and maps them to corresponding form fields.

Google emphasizes that sensitive data accessed by Enhanced Autofill remains encrypted and that the user retains control: individual suggestions can be removed, the feature can be completely disabled, or permissions managed directly in Chrome settings. The feature requires the user to have Personal Intelligence enabled — an additional opt-in layer that already separates the feature's adoption from Chrome's base installation.

In the current stable version, access is limited to Gmail. But Google's team has explicitly mentioned the intention to expand to Google Photos (where Chrome could extract information from saved images) and potentially to other connected apps, following the model already established by Gemini itself.

Google's glue strategy

This update doesn't exist in a vacuum. It's the most visible piece of a long campaign to transform Chrome from a simple browser into a personal productivity platform. Google has invested billions integrating Gemini into every corner of Chrome — with the side panel for multitasking, auto browse for autonomous navigation, generative AI search — and now autofill is another way to glue the user to the ecosystem.

The Apple analogy is unavoidable. Apple has long consolidated the most praised feature of iOS: automatic detection of verification codes sent via SMS, displayed above the keyboard for one-tap entry. Chrome's Enhanced Autofill is Google's desktop/mobile equivalent — less magical in the "one-tap" sense, but potentially more transformative in terms of time saved.

The timing is revealing: Chrome is under pressure. Firefox with its AI smart windows, Opera with the WebAI initiative for local model execution, and the growing popularity of alternative browsers like Zen (the Arc successor, discontinued in December 2026) have created a more competitive environment than seen five years ago. AI-powered autofill is one way to defend Chrome's market share with a functionality that competitors still don't offer at comparable maturity.

The privacy question — and why you should care

Here's where the real friction lies. Google knows you open hundreds of emails per month. Some contain financial information, passwords, health data. Saying that Gemini "can read your Gmail" — even in encrypted form, even on-demand, even with an opt-out — is a statement that should make privacy-conscious users uncomfortable.

The Android Authority correctly pointed out that Google recently updated its privacy policy to allow the use of Personal Intelligence data for ad personalization. This doesn't mean Enhanced Autofill will use Gmail data for advertising — Google claims to keep the two streams separate — but it creates a concerning precedent: if Personal Intelligence can be used for ads, Enhanced Autofill is the first bridge into that flow.

Additionally, there's a more subtle question: what is the attack surface of an AI model that scans your email for data to autofill? Even if encrypted, the process requires Gemini to read Gmail content. If the model runs locally, the risk is lower. If it depends on cloud inference, each autofill request could be a potential exposure point. Google has not disclosed details about where exactly the processing occurs.

What to watch in the coming months

Three questions will determine whether Enhanced Autofill becomes an adopted feature or a curiosity that no one bothers to disable:

1. The accuracy of suggestions. If Gemini suggests the correct tracking number, it's a win. If it starts filling the "name" field with an order number because the model confused the context, frustration will be immediate. Trust in autofill is built on absolute precision — one failure destroys that trust.

2. The expansion to Google Photos. Google has already mentioned that the next expansion will come from images stored in Google Photos — meaning Chrome could extract information from screenshots, digitized receipts, and documents. This dramatically expands both the utility and the privacy risk.

3. The competitor response. Firefox Smart Window already offers AI form filling via Gemini and Qwen. If Mozilla or Opera respond with a comparable (or better) offering, Enhanced Autofill loses its monopoly on differentiation. If they lag, Chrome consolidates another choke point in the ecosystem.

Chrome's autofill has never been the most glamorous story in technology — but it's one of the most frequent interactions billions of people have with a browser, daily. When Gemini enters that flow, what's at stake isn't just convenience: it's the definition of who controls your personal data at the most mundane and revealing act of web browsing.

Sources: Google Blog, TechCrunch, The Verge, Android Authority

✓ Independent sources cross-checked and verified before publishing