When someone says they're building an "AI game" in Unreal Engine, they could mean one of two completely different things, and the confusion trips up a lot of newer developers. So before any tutorial, the most useful thing I can give you is the distinction:
Classic game AI — the decades-old discipline of making NPCs behave: pathfinding, decision-making, enemies that flank you. This is deterministic, authored, and shipped in basically every game you've played.
Generative AI — the new wave: large language models giving NPCs open-ended conversation, plus AI tools that generate worlds and assets.
They share two letters and almost nothing else. Let's go through both as they stand in Unreal in 2026, because a strong game usually needs the first and is starting to experiment with the second.
Part 1: Classic game AI (the kind that ships today)
