FRENES

Glossary · Everyday AI use

Browser extension

A browser extension is a component you add to a web browser to provide extra features or change certain behaviours. It might help you summarise a page or organise information, depending on the permissions you grant, the browser’s capabilities and any services the extension uses.

Browser extension: the explanation

An extension uses facilities provided by the browser to add an interface or interact with pages within its permissions. The [Mozilla documentation](https://developer.mozilla.org/en-US/docs/Mozilla/Add-ons/WebExtensions/What_are_WebExtensions) explains this approach to extending browser functionality. You might see a toolbar button, a side panel or an additional menu action. Some extensions operate only when invoked, while others can work more broadly on permitted sites. Their actual behaviour depends on their design, browser settings and the permissions you have granted rather than on their visible interface alone.

An AI extension often brings assistance directly into your normal browsing activity. You might request an article summary, an explanation of a passage or help drafting text in a page. The extension may process a selected fragment or a larger body of content, depending on how it works. It may also contact a remote service. Seeing its controls inside the browser therefore does not establish where processing takes place or which information is sent away to produce the requested response.

Permissions should be assessed against the extension’s actual purpose. A tool intended for selected sites may not need access across all your browsing. Available restrictions vary between browsers and extensions, however. Web pages can also contain misleading instructions aimed at an agent reading their content. If an extension can act on your behalf, inspect the confirmations and limits surrounding those actions as well as its data access. Reading a page and changing information in an account are different capabilities with different consequences.

A concrete example

You install an extension to summarise articles. On a public page, you select a passage and request a summary. Before using it in your work email, you inspect its permissions and data-handling information. You check whether it sends only selected text or can access additional content. Where the browser and extension allow it, you restrict the sites on which it can operate before using it with more sensitive information.

Why it matters

An extension places AI features close to the pages you visit, sometimes with substantial access. When choosing one, examine its source, permissions and data transfers. Check whether it only reads information or can also act inside your accounts. A small interface may provide broad capabilities, so focus on what the extension can inspect, transmit and modify during browsing rather than judging it by its visible size.

Common misconceptions

  • An extension installed in your browser does not necessarily process information on your device, because it may send text or other data to a remote service to generate a response according to the way it was designed.
  • An extension appearing in a catalogue is not a guarantee that it suits every use, because you still need to assess its permissions, data-handling practices and any ability it has to perform actions inside your accounts.

Related terms: everyday ai use

prompt
A prompt is input supplied to a model to guide what it should produce or do.
RAG (retrieval-augmented generation)
RAG, or retrieval-augmented generation, combines information retrieval with a model’s ability to generate an answer.
vector database
A vector database stores and searches numerical representations called vectors.
semantic search
Semantic search retrieves material according to its likely meaning, beyond matching the same words.
coding assistant
A coding assistant is an AI tool that helps you write, understand, change or test computer programs.
no-code
No-code is an approach to building applications or automations through visual interfaces and settings, without directly writing code for the supported functions.
text-to-speech
Text-to-speech is a technology that converts written text into machine-generated speech.
speech recognition
Speech recognition is a technology that turns spoken audio into text a computer can process.

Where this site covers it

The term appears on 169 pages of this site. A few of them:

Same family

Go further

Frequently asked questions

Can it see every page?

That depends on its permissions, implementation and browser settings. Some extensions can be restricted to particular sites or individual interactions. Review the effective access it has rather than inferring its capabilities from the button or panel it displays.

Does it work offline?

Some features can operate locally, while others require a remote service. This depends on the extension and the model it uses. Check its documentation and settings: installation inside the browser does not mean that every function works without a connection.

Can you restrict its access?

Your browser may offer site-specific or on-demand permissions, but the available controls vary. Review those settings alongside the extension’s requirements. Remove unnecessary access where possible, then check that the functions you need still work with the restrictions applied.

Neutrality. No affiliate links. No vendor pays to be reviewed, rated or ranked. Reviews are directed by François Chollet, neutral and factual, and every data point is dated and sourced. How we score