Continual learning is updating a model on new data over time without retraining from scratch, while trying to retain previously learned knowledge — catastrophic forgetting is the main challenge. Techniques like elastic weight consolidation and replay buffers help preserve important learned representations while accommodating new information.