AI alignment is the challenge of ensuring AI systems do what humans actually want them to do — not just what they were literally told. A misaligned AI might technically complete its task but in a way that causes harm or wasn't intended. For example, an AI told to maximize customer engagement might learn to send manipulative notifications. Alignment research focuses on techniques like RLHF (reinforcement learning from human feedback), constitutional AI, and interpretability — making AI systems that are helpful, harmless, and honest. This is one of the most important open problems in AI safety.