The “where she at” AI video, explained

In the original clip from Drake’s “Burning Bridges” music video, the lead walks through a dim, chandelier-lit dinner party, holds a hand over his eyes and scans the room as if asking “where she at?”

A hand held flat over the eyes is the look-out pose: the gesture of someone searching a crowd. Paired with a slow look around a party, it reads as a question without any words. “Where she at?” is how viewers put that question into text.

That is all the phrase needs to mean here. You do not have to know the rest of the song or the video to get the joke; the gesture carries it, and the caption names it.

In late September and early October 2026, AI versions of the party scene spread across TikTok, X and Instagram Reels, with people, friends and even pets swapped in as the lead.

A meme needs a short label people can type, search and repeat. The scene has no dialogue to quote, so posters used what the gesture looks like: someone asking where she is. The phrase works on its own and leaves room for a setup: who is “she”, and who is looking for her?

That is why searches like “Drake where she at” and “where she at AI video” lead to the same clips as “Burning Bridges AI video”. They are two names for one trend. For the full background, read what the Burning Bridges AI trend is.

You make it the same way as any Burning Bridges AI video: the party scene supplies the movement, and a photo supplies the face.

• The one-photo way: upload a clear photo of one person on the home page, sign in with Google and use 1 credit. You get a 720p clip of about 10 seconds with the song audio, usually in a few minutes.
• The do-it-yourself way: a motion-transfer tool such as Higgsfield Genjutsu or Kling motion control, with the party scene as the driving clip and your photo as the reference. Step by step in how to make a Burning Bridges AI video.
• The phone template way: CapCut templates built around the trend. Quick, but the quality varies by template and templates get removed.

Whichever you pick, a sharp, front-facing photo with the whole face visible makes the biggest difference. The photo guide explains why, and how to take one in two minutes.

The format is a setup and a punchline: the caption asks the question, and the person scanning the room is the joke. Some ways to use it:

• Reaction post. Pair the clip with a situation everyone knows: looking for the one friend you know at a party, or for the waiter after a long wait.
• Birthday. Put the birthday person in the lead, with their OK, and post it on the day with “where the cake at?”.
• Group chat. Send it when someone has gone quiet, with that friend as the lead. Ask them before it goes beyond the chat.
• Waiting on someone. Put yourself in the lead and caption it about who you are waiting for. Keep yourself as the lead rather than someone who did not agree.
• Work and school. You, scanning the room for your group project partner the day before the deadline.
• Pets. People have made versions with dogs and cats using general motion-transfer tools. Burning Bridges AI Video does not take pets: it puts one person in the lead role, so it needs a photo of one person.

The common thread: the caption sets up who is missing, and the clip shows who is looking. Keep the setup to one line so it reads before the hand comes up.

• Put the setup in the caption or as on-screen text, and let the clip be the punchline.
• Use the official “Burning Bridges” sound from the TikTok or Reels sound library so your post appears on the sound’s page with the rest of the trend.
• Post vertical (9:16) for TikTok and Reels, and keep the hand gesture inside the frame if you crop.
• Keep it short. The clip is about 10 seconds and loops; an intro before it only loses viewers.
• Add #wheresheat next to #burningbridges so people searching either name find it.
• Label it as AI where the platform asks, and only tag people who agreed to be in it.

Need a caption and hashtag set to start from? The prompt guide has one you can copy.