The principle that AI systems should treat all individuals and groups equitably, without bias, discrimination, or favoritism. This involves ensuring that the outcomes produced by AI models do not disproportionately harm or benefit any particular demographic, and that the system?s design, data, and implementation processes actively prevent unjust or prejudicial treatment.
