Artificial intelligence is trained on real intelligence – in other words, content created by humans. Theoretically, if you want AI to spit out fiction, you train it on novels and if you want AI to write job specs, you train it on job descriptions written by people. But there’s a problem here. No matter how hard humans try to be unbiased, we all have some. And an AI that has been trained on human work will take on some of these human biases. Indeed, overcoming biases is one of the biggest challenges in AI.
“In recruitment, bias is everywhere and comes from a baseline of most recruitment processes being run by humans,” says Kevin Fitzgerald, UK managing director of Employment Hero, a global employment management platform that is using AI to make doing business easier for SMEs. “The underlying bias that recruitment teams have is there in the process already.”
