Organisations need to adopt basic principles for responsible artificial intelligence to gain competitive advantage; they must also design systems that can effectively deal with biases