Abstract: We will all be aware of algorithms that, when let loose in the real world, have produced bad outcomes. One of these was recently described by the British Prime Minister as 'mutant', a description that was criticised as downplaying the algorithm's design and the responsibility of the humans that deployed it. Individuals and regulators can use the law to challenge unfair outcomes brought about by 'mutant' algorithms. But this talk will suggest that more can be done to ensure that the principles of key legal frameworks are built into the design of algorithm-assisted state decision-making and in the consideration of the consequences of operational deployment. In doing so, independent and rolling review can play a significant role.