ALGORITHMIC BIAS
ALGORITHMIC BIAS
Algorithmic bias can occur if:
the data used to train AI is too narrow (is not representative of the whole population)
the training data is historical and doesn't reflect current trends and values
the training data is simply too small
the team of programmers are themselves introducing bias
there is insufficient human oversight on how the algorithms are learning
Watch/listen to the presentation/video on the right. It is based on a 2025 article.