Artificial intelligence (AI) systems are becoming increasingly complex to solve various problems more efficiently and accurately. Consequently, AI explainability has become crucial for ensuring trust in AI. This chapter reviews different concepts and methods of AI explainability, highlighting the challenges faced by AI explanations. Since humans are the primary end users of most AI applications, incorporating a human perspective into AI explainability is essential. This chapter introduces a human-centered approach to the AI lifecycle, linking it with AI ethical theories, and suggests that human-centered explanations are necessary at every stage of the AI lifecycle. Focusing on the AI model-building stage, this chapter proposes a human-centered explainable model framework, discusses the challenges of human-centered AI explanations, and provides research directions for future development in this area.