Last week, New Jersey Attorney General Matthew Platkin announced new guidance that the New Jersey Law Against Discrimination (LAD) applies to algorithmic discrimination, i.e., when automated systems treat people differently or negatively based on protected characteristics. This can happen with algorithms trained on biased data or with systems designed with biases in mind. LAD prohibits discrimination based on a protected characteristic like race, religion, national origin, sex, pregnancy, and gender identity, among other things. According to the guidance, employers, housing providers, and places of public accommodation who make discriminatory decisions using automated decision-making tools, like artificial intelligence (AI), would violate LAD. LAD is not an intent-based statute. Therefore, a party can violate LAD even if it uses an automated decision-maker with no intent to discriminate or uses a discriminatory algorithm developed by a third party. The guidance does not create any new rights or obligations. However, in noting that the law covers automated decision-making, the guidance encourages companies to carefully design, test, and evaluate any AI system they seek to employ to help avoid producing discriminatory impacts.
[View source.]
