An early AI was modeled on a psychopath. Researchers say biased algorithms are still a major issue

It started as an April Fools' Day prank.

On April 1, 2018, researchers from the Massachusetts Institute of Technology (MIT) Media Lab, in the United States, unleashed an artificial intelligence (AI) named Norman.

Within months Norman, named for the murderous hotel owner in Robert Bloch's — and Alfred Hitchcock's — Psycho, began making headlines as the world's first "psychopath AI." 

But Pinar Yanardag and her colleagues at MIT hadn't built Norman to spark global panic.
It was supposed to be an experiment designed to show one of AI's most pressing issues: how biased training data can affect the technology's output.