There is increasing concern that machine learning tools embed bias in their operations and outputs. To what extent does the law currently provide adequate protection from or adequate redress in respect of any such discrimination?