Stop Building Your AI Video Stack Before You Do This One Thing First

Your client just rejected the AI-generated b-roll. They can’t tell you exactly why — it "just looks off."

You’ve checked the resolution, the lighting matches the brief, the motion is smooth. So what’s actually happening?

After working through dozens of AI video generations for commercial use, the rejection almost always traces back to four specific tells. None of them are about resolution or polish. They’re about physics and attention.

1. Motion that’s too consistent

Real cameras have micro-variance. A handheld shot drifts slightly even when "locked." A drone shot has subtle wind correction. AI-generated motion is often too smooth — perfectly even acceleration, perfectly stable framing. Our eyes are tuned to detect that kind of unnatural consistency, even if we can’t name it. It reads as synthetic before a single frame is consciously analyzed.

2. Lighting that doesn't argue with itself

In real footage, light sources compete. A window throws cool light while a practical lamp throws warm light, and they fight for dominance across a face or surface. AI models tend to harmonize lighting into a single coherent mood — which looks good, but real environments are rarely that resolved. The harmony itself becomes the giveaway.

3. Texture that’s clean in the wrong places

Skin, fabric, and worn surfaces should show inconsistency — pores, fraying, scuffs. AI video tends to either over-smooth these details or apply texture too uniformly across a surface, so a sleeve and a countertop end up with oddly similar material logic. It’s not that the texture is wrong, it’s that it’s too agreeable.

4. Background elements that don't have independent lives

In real footage, background elements move on their own physics — a flag in the wind, blinking extras, distant traffic. AI-generated backgrounds often move in service of the foreground composition rather than independently. The background feels staged because it technically is, in a way that real footage rarely is.

What this means for your workflow

None of these are fixable by just generating more takes or tweaking the prompt for "more realistic." They’re structural to how these models currently render scenes. The actual fix is selective: use AI footage for shots where these tells are less visible — wide establishing shots, abstract texture inserts, motion graphics-adjacent content — and reserve real footage or hybrid compositing for close-ups, dialogue-adjacent b-roll, and anything with skin or fabric in focus.

Clients don’t need the technical explanation. They need you to know, before they see the cut, which shots will survive scrutiny and which won’t.