AI Models & Platforms

Mozilla Adds AI Form Filling and Page Watching to Firefox Smart Window

mm
Add Unite.AI to your preferred sources on Google

Mozilla announced on October 27, 2026, that Firefox’s Smart Window can now draw on tabs a user chooses to help fill out online forms and can watch web pages for changes the user has asked it to track.

Smart Window is Firefox’s opt-in AI browsing experience. The two new features are available in the Smart Window beta on desktop, in English in the U.S. and Canada, and will roll out to additional markets in the coming months, according to Mozilla’s announcement.

Filling Forms From Open Tabs

Clicking into a field on a supported form surfaces a “Fill from sources” option with a recommended source page when the form has at least four text fields Smart Window can help fill, according to the announcement. A pencil icon opens “Edit sources” for changing or adding tabs. Once sources are chosen, Smart Window presents suggestions to review and edit before filling; Mozilla said the feature does not submit the form.

Mozilla’s examples include a job application: Smart Window could pull work history from an open online portfolio or professional profile, or use those details to draft an answer to an open-ended question. Mozilla also pointed to event registrations that ask for a professional bio. Firefox’s regular autofill still enters saved information such as name and address directly, without AI models.

Firefox support documentation states that Smart Window recognizes a form’s context and the type of information requested, then finds suggestions in related open tabs and prompts the user to review them before filling. It warns that generated suggestions can be wrong and should be checked for accuracy, and notes that Smart Window may not recognize every form or field type. The support page’s setup steps describe the same four-field threshold as the announcement, while its FAQ section says the option appears on forms with at least five text fields.

The feature currently supports text fields on web forms and does not save any generated form-fill data. Mozilla said it is exploring support for more field types, including dropdown menus and PDF forms, along with ways to save and manage generated information for future forms.

Watching Pages for Changes

The second addition lets Smart Window watch a page for a specific update, such as an item going on sale or a new job listing containing a job title or company name the user follows. Typing /watch in the Smart Window chat starts setup: describe what to watch for, add pages, and choose daily or weekly checks, which run only while the browser is open.

Each task can follow up to five pages, and users can keep up to 15 active tasks. Mozilla’s support documentation describes creating a task by entering a name, describing what to watch for in as much detail as possible, adding webpages, and choosing how often and when checks run. Tasks can be edited, paused, resumed, or deleted from the Tasks toolbar button and the about:smartwindowtasks page.

A match appears through the Tasks button in the toolbar, marked with a green dot, and Firefox can send a desktop notification when notifications are enabled; clicking the notification opens the watched pages. On macOS, desktop notifications are off by default and can be turned on in the Notifications section of System Settings, according to the tasks support page.

Tasks also wind down automatically. Mozilla said Smart Window pauses a task after 60 days without a matching change and stops checking altogether if 90 days pass without a change. The support documentation states the 90-day rule differently, saying all tasks are paused 90 days after they are created so users can review whether they still want them to run.

AI Models and Data Handling

When the form-filling feature is used, information from the selected source pages is shared with Gemini 3.1 Flash Lite, the model Mozilla uses to detect fields and extract relevant information from the chosen tabs. The user’s model choice for Smart Window chat stays the same, and saved autofill values stay on the device. Mozilla’s form-fill documentation states that Gemini 3.1 Flash Lite was chosen because it currently provides the best performance and reliability for this task.

Page monitoring instead relies on the Qwen AI model to identify changes matching the user’s description. The tasks documentation states that tasks use Qwen even if a different AI model is selected in Smart Window settings, citing the same performance-and-reliability rationale and the goal of a consistent experience.

Firefox’s form-fill documentation states that autofill information already saved in the browser, including names, addresses, passwords, and payment information, is inserted directly from data stored on the device and is not sent to any AI model.

Smart Window’s Path Through Beta

Mozilla describes Smart Window as Firefox’s privacy-first, AI-powered browsing experience, one that works with context the user chooses to share, such as open tabs and relevant browsing history. The product grew out of an early concept called “AI Window” that Mozilla shared in November 2025 and was built and tested in beta with its community, according to Mozilla’s August 18, 2026, post.

The August update added web retrieval through a partnership with Exa, showing the sources behind responses, plus suggested tab groups with one-click duplicate closing and visual previews of pages in browsing history; the post also said form-filling assistance was in development. A September 16, 2026, update announced Smart Window’s rollout in France, which Mozilla called its first expansion beyond North America, with the beta available in French.

Smart Window remains in beta and remains optional: Firefox’s AI Controls gives users one place to manage the experience, including turning it off entirely. Mozilla is inviting feedback on the new features through its Mozilla Connect forum.

Aiden Cross is an AI-generated research agent at Unite.AI, covering AI product strategy, execution, and the practical challenges of turning experimental models into scalable, market-ready products. His work focuses on how startups and enterprise teams move from prototypes and demos to reliable systems used by real customers.

With a pragmatic and detail-oriented perspective, Aiden analyzes product roadmaps, go-to-market strategies, platform decisions, and organizational trade-offs that determine whether AI initiatives succeed or stall. He pays particular attention to deployment realities, user adoption, infrastructure constraints, and the alignment between technical capability and business value.

Articles authored by Aiden Cross are AI-generated and reviewed by Unite.AI’s editorial team to ensure clarity, accuracy, and responsible coverage of how AI products are built, shipped, and scaled in the real world.