Nowadays, more people get highly interested in AI's unethical and biased settings as it becomes part of society. This paper aims to define AI and emphasizes that explainability, transparency, responsibility, non-discrimination, data protection, and good intention should guide practitioners and researchers when designing and implementing AI systems. This section breaks AI bias down into data bias, algorithm bias, and social prejudice to explain it. This section also provides working examples of different AI prejudice systems to demonstrate their influence. Discrimination preferences, bias, privacy, and decision making are thoroughly examined as AI ethical challenges. Pre-processing methods emphasize the use of diversely obtained data, appropriate and accurate bias identification and elimination measures, and the integration of bias mitigation strategies at the design stage to avoid and handle bias. The paper provides actionable principles and recommendations for AI developers, policymakers, business, academia, and civil society on how to develop and utilize AI technologies in the public interest soon. They analyze future AI ethics research trends and propose policy recommendations for policymakers and AI practitioners to develop ethical AI that benefits all individuals and societies.