To make AI fair, here's what we must learn to do

Beginning in 2013, the Dutch government used an algorithm to wreak havoc in the lives of 25,000 parents. The software was meant to predict which people were most likely to commit childcare-benefit fraud, but the government did not wait for proof before penalizing families and demanding that they pay back years of allowances. Families were flagged on the basis of'risk factors' such as having a low income or dual nationality. As a result, tens of thousands were needlessly impoverished, and more than 1,000 children were placed in foster care. From New York City to California and the European Union, many artificial intelligence (AI) regulations are in the works.