Synthetic data is artificially generated information that statistically mirrors real-world datasets without containing actual personal or sensitive information. It’s increasingly used to train AI models when real data is scarce, sensitive, or biased.