Lightroom Classic now has AI-assisted culling that can scan your entire shoot and tell you which frames actually meet your criteria. When you're staring at rows of nearly identical portraits after a long session, this kind of help saves you from spending hours doing basic sorting.
This clear video from Aaron Nace and Phlearn explains how AI Assisted Culling connects directly to the loading screen in Lightroom Classic. Nace starts by dragging a folder of sample images into the app and turning on the new assistance option, allowing the software to analyze them all at once. The AI checks whether the subject is in focus, the eyes are in focus, and whether the eyes are open and marks each file with a green check mark or red X, so you can immediately see which shots passed the technical test. You'll also see that the panel automatically rejects documents, receipts, skimpy frames, etc., so clutter and misfires don't even get cataloged. The whole idea is to at least fill the catalog with only usable images, rather than every frame you happen to fire at.
Nace then shows how to take this further than a simple yes or no filter. In the import dialog, you can drag a slider that controls how precise the “subject in focus” actually is. This allows you to decide whether a focus score of 81 is acceptable or whether only the absolute sharpest shots will pass. If you want to keep more options, where facial expressions and moments can be more important than pixel perfection, you can loosen those settings and include slightly softer images. The video also makes clear that the AI does not override your preferences. If the software doesn't agree, you can still turn thumbnails on or off manually. The balance of automation and control is what turns this from just a gimmick to something you can actually rely on for everyday editing.
Once the imports are set up, Nace moves on to how assisted culling works within library modules. The same focus and wide-eyed checking is available there, alongside stacking and batch tools. Group images can be automatically stacked by capture time, so individual shots on a single memory card are categorized into their own clusters. Visual similarities can also stack up in long bursts of nearly identical poses. Clicking on the stack number will expand the group, and if you want to compare the frames in more detail, you can use the menu options to expand, collapse, or unstack everything. From there, Nace uses batch actions to apply white flags to selected keepers and black reject flags to misses. If you need more structure, use the optional star ratings and color labels. You can still see moments where he overrides the AI on a case-by-case basis, which is useful if you prefer images with slightly lower focus but higher energy.
This video also explains how to actually access your favorites after the AI has done its job. Nace displays the library filter bar, switches to the attributes section, and filters by flags so you only see the images you selected or rejected when making a decision. Then, use Survey view to compare a small set of similar portraits side by side, removing weak frames one by one until only one favorite remains, and assigning a 5-star rating to that final selection. Keyboard shortcuts such as N in the survey view, the slash key in the filter bar, and the number key for ratings can speed up the process without adding complexity. You can also get a glimpse of how this culling step connects to AI-driven masking, denoising, lens blur, and color tools in Lightroom Classic and Photoshop in later edits. Watch the video above for Nace's complete overview.
