anti-discrimination laws

Terms from Artificial Intelligence: humans at the heart of algorithms

Page numbers are for draft copy at present; they will be replaced with correct numbers when final book is formatted. Chapter numbers are correct and will not change now.

Many nations have laws to prohibit particular forms of descriminiation often including gender, ethnicity and sexual orientation. For example, the UK list of protected characteristics at the time of writing includes age, gender reassignment, being married or in a civil partnership, being pregnant or on maternity leave, disability, race including colour, nationality, ethnic or national origin, religion or belief, sex, and sexual orientation. The laws may have severe financial or other penalties for companies who have been found to have acted in a discrimnatory manner, for example, having software that has some form of bias. This is parricularly probematic for compex AI algorithms and machine learning such as deep neural networks as it may be unclear what aspects are being used to make a decision. Note that irrespectuve of the law in the nation(s) in which you are working or deploying systems, you should consider the ethical consequences of systems you produce, for example, bias for non-protected characteristcs such social background, or wider societal impact of AI.

Used in Chap. 20: page 335