Humans are ready to take advantage of benevolent AI

Humans expect that AI is benevolent and trustworthy. A new study reveals that at the same time humans are unwilling to cooperate and compromise with machines. They even exploit them.