Datafication, Artificial Intelligence, and Rule of Law

Artificial Intelligence has been touted as a technology that would significantly contribute to the rule of law and the achievement of Sustainable Development Goals, such as SDG 16. AI-based systems are presented as doing away with the messiness and inconsistency of human judgement due to their deterministic structure, or as uniquely responsive to societal needs through Big Data analytics, and as granular and real-time in nature, i.e. as capable of adjusting regulation to an individual&s;s personal circumstances. However, these same features simultaneously run against some of the values crucial to both rule of law and egal access to justice: real-time granularity challenges legal stability and equality before the law while also making it impossible to know in advance the law one is subjected to; machine learning techniques applied to sets of Big Data are bound to learn from and strengthen the biases and inequalities present in a society; while the determinism of computer systems will lead to mindless application of whatever is learned, without due regard to the broader context. Moreover, the growing power of AI severely disrupts the balance between public and private governance, rendering it a stumbling block in the further realisation of SDG 16.