Leaked Pixel 8 video shows off Google's next AI camera features

Getting a group of people to all smile and look at the camera at the exact same moment is practically a statistical impossibility. Google’s new answer? Just don't bother, and let AI stitch a fake reality together for you instead.

A major leak of the upcoming Google Pixel 8 line just dropped, courtesy of leaker Kamila Wojciechowska and 91Mobiles. Along with hardware specs, a leaked promotional video reveals Google’s next wave of aggressive AI photography features.

The Frankenstein group shot

The marquee feature shown off in the video is Magic Editor, which Google first teased back at I/O in May. It plays out like an absolute supercharged version of Magic Eraser.

In one demo, someone snaps three rapid shots of a family on a carousel, each photo flawed by someone looking away or frowning. Magic Editor takes the best expressions from the three frames and blends them into one single, impossible photo where everyone is smiling directly at the lens.

You can also drag and reposition people around the frame, remove background objects entirely, or swap a midday sky out for a sunset. When you change the sky, the lighting across the whole shot recalculates to match the golden hour vibe.

For traditionalists, Google is also finally adding DSLR-style manual controls. Instead of the typical tap-to-focus interface, you get dedicated sliders for shutter speed, ISO, and manual focus.

The hardware upgrades

The leak also detailed what is under the hood for both camera modules.

The Pixel 8 Pro gets the real hardware love: a 50MP main wide shooter, flanked by an upgraded 48MP ultrawide with a 125.5-degree field of view and a 48MP telephoto.

The vanilla Pixel 8 shares the same 50MP main sensor. However, its ultrawide stays at 12MP, though Google expanded its field of view from 114 degrees on the Pixel 7 to 125.8 degrees.

Why it matters

For everyday users, this feature will save countless ruined group photos and vacation snaps. Anyone with kids or large families knows the pain of chasing the perfect frame. Being able to effortlessly fuse three mediocre shots into one clean memory is undeniably useful.

Yet it pushes mobile photography right off the cliff into ethical ambiguity. We have spent decades accepting photo editing, but merging multiple moments into an event that literally never occurred turns photography from documentation into creative fiction.

Industry-wise, we think Google is playing this very smartly. Instead of marketing abstract chat agents, it is embedding generative AI directly into the primary reason people buy smartphones: the camera. This will force Apple and Samsung to accelerate their own generative on-device editing pipelines.

The risk, of course, is uncanny valley territory. If hair cutouts look fuzzy or lighting adjustments clash with shadows, photos will quickly look synthetic and hollow. How well it handles these edge cases in the real world will decide if it is a genuine daily tool or just a party trick.