Artificial Intelligence (AI) has the potential to build upon the capabilities, experiences, and insights of human beings, transforming how we approach various tasks and challenges by supplementing human-like abilities. The process of augmenting human intelligence refers to the use of AI to enhance human intelligence, rather than operating independently of, or replacing it. This use of AI is designed to include and balance human oversight across the AI lifecycle. Thomas Watson Jr., IBM’s 2nd President, once said, “Our machines should be nothing more than tools for extending the powers of the human beings who use them,” and we continue to stand by this point of view today.

IBM’s Principles for Trust and Transparency delineate the ways in which IBM embeds trust into its technology solutions. The first of these principles is: the purpose of AI is to augment human intelligence. All AI systems developed and deployed by IBM are designed to augment, not replace, human intelligence. IBM achieves this by designing such systems to enhance and extend human capability and potential while simultaneously enabling humans to take on more value-adding tasks. Accordingly, AI should improve job performance and augment individual as well as business performance. This implies that AI systems are to act as support functions to enhance human intelligence, rather than operate as completely separate, autonomous entities.

AI should be designed to include and balance human oversight across the AI lifecycle: when operations are carried out by AI, humans, or humans augmented by AI. This further supports IBM’s ability to embed ethical governance into developing, deploying, and building AI systems that help clients succeed.