Privacy‐Preserving AI Techniques

AI is transforming business, including the medical, banking, and education industries, because it finds better solutions through the analysis of huge amounts of data. Regardless of this fact, the use of data triggers negative issues with privacy, especially in the context of confidential information. Growing concern over data security and privacy breaches has sparked increasing regulation with the likes of GDPR and CCPA. The chapter presents the overview of AI techniques that guarantee privacy including differential privacy, federated learning, homomorphic encryption and safe multi-party computation. These approaches allow AI to give results while anonymizing the data. This competition highlights real-life scenarios which illustrate how these approaches are revolutionizing the prospects of AI. If we start from a principle where privacy is the dominant characteristic of AI then AI will be moral, trustworthy, advantageous for humanity and global society, and will not be in contradiction with the individual.