A video that follows you while you read another page can be genuinely useful. A lecture stays in view while you take notes; a tutorial remains visible beside the task it explains. Picture-in-Picture, or PiP, is the browser feature that puts a video into a small floating window. Its convenience also raises fair questions: who controls that window, when can a site create one, and what can the site learn from it?
What the Picture-in-Picture API does
The standard video Picture-in-Picture API lets a page ask the browser to display one of its video elements in a floating player. The player can sit above other windows while you move to another tab or application. The video still belongs to its original page: PiP is a different presentation of that media, not a download or a new copy of the video.
MDN’s API guide describes the page’s request method, HTMLVideoElement.requestPictureInPicture(), and the document method that exits the mode. It also describes events for entering and leaving PiP. A site can use those events to update a button or pause an interface animation; the existence of a floating player does not grant it a view of every window on your desktop.
Some browsers also offer their own PiP control on a video, independently of a site’s custom button. The exact menu, keyboard shortcut, window controls, and supported media vary by browser and platform. Look for a small floating-window icon in the player or browser controls, then use the close or return-to-tab control when you are done.
Why a page cannot simply pop up video at will
Opening a floating window without warning would be distracting. The video PiP request is among web features that require a recent user interaction, such as clicking a control, according to MDN’s user-activation reference. A site generally cannot load quietly in a background tab and use this method to put a player on top of your work. Browser-specific automatic PiP experiences can have separate rules, so the behavior you see may not always come from a page calling this API.
The browser can reject a request if PiP is unavailable, the element is unsuitable, or policy disallows it. The picture-in-picture Permissions Policy lets a site restrict whether documents, including embedded content, may use the feature. A web developer can also set a hint that PiP should not be suggested for a particular video. These are boundaries around the experience, not a promise that all browsers will present identical controls.
What a website can and cannot tell
The originating page can know that its own video entered or left PiP and can receive resize events for the PiP window. That is necessary to keep its player interface in sync. It does not receive an unrestricted list of your other applications or a recording of your desktop through this API. PiP is therefore not the same as screen sharing. Our screen-sharing explainer covers the separate, explicit selection involved when a site requests a screen or window capture.
There is a different practical privacy issue: the floating video may remain visible to people nearby after you leave its original tab. If a lesson, meeting, or personal video contains sensitive material, close PiP before sharing your screen, stepping away, or inviting someone to look at your desktop. A small window is still a window. The browser cannot infer which video you consider private.
Also remember that the originating site can continue ordinary page activity while its tab remains open. PiP does not itself switch off analytics, cookies, or network requests. If you want to stop the site’s activity, close the tab or use the browser’s normal privacy controls, rather than treating the PiP close button as a complete site shutdown.
Video PiP versus document PiP
The familiar version floats a video. A newer, separate Document Picture-in-Picture API can place a small HTML document in an always-on-top window on supporting browsers. That can be useful for custom controls or non-video tasks, but it is not the same capability or compatibility story. It also has an interaction requirement. If a site shows a complex floating panel instead of a plain video, look at which browser feature it is using before drawing conclusions from the standard video API.
For users, the distinction is simple: ask what content is in the floating window and whether you deliberately opened it. For developers, it means providing clear controls, a useful fallback when PiP is unavailable, and a sensible way to return to the original page. A feature that works only when a user can find the exit button is not a thoughtful feature.
A practical way to use it
Start PiP from a video you chose to watch, resize and move it to a place that does not cover important work, and close it when the context changes. If a site tries to push an unexpected floating experience, leave the page. If a video keeps playing after you close PiP, return to the source tab and pause or close it there. The PiP window and playback state are related, but they are not always the same control.
Our Page Visibility API guide explains what pages can learn when you switch tabs; our permissions guide helps distinguish display conveniences from genuine data access. Picture-in-Picture works best when that boundary is obvious: your choice to keep one video visible, without implying that the site can see the rest of your screen.
Browse with more intention
Noorani brings prayer times, Qibla, tracker blocking, and privacy into one calm desktop browser built for how Muslims live online.