“Artificial Intelligence, Machine Learning and Deep Learning – Limitations: Privacy and Data Security Issues” Chapter

Abstract

The opportunity to develop amazing AI tools to improve and reimagine healthcare is exciting and being explored by many companies and startups alike. The adoption of AI in daily clinical practice to reduce cost and potentially improve quality and accessibility although urgently needed is not guaranteed. Constraints exist on data - a main element required to develop such AI tools. Data that is collected, processed, enriched and retained to train and validate AI models is subject to privacy and security laws and regulations. These laws and regulations are progressively being strengthened. In addition, there are ethical frameworks starting to be formulated that must be considered when developing and using AI tools that guide best practice and behaviour. These two areas are important elements to engendering trust in the development and use of beneficial AI technology in healthcare. In this chapter we provide a summary of these issues with some examples that aim to help anyone wishing to build and deploy AI tools in healthcare to do that in a responsible, legally compliant and ethical manner.