A/HRC/44/57 targeted. In other words, the data is more predictive of racialized disadvantage and police presence in an accused person’s community than a person’s behaviour.”98 IV. A structural and intersectional human rights law approach to racial discrimination in the design and use of emerging digital technologies: analysis and recommendations 44. Emerging digital technologies pose a mammoth regulatory and governance challenge from a human rights perspective. In many cases, the data, codes and systems responsible for discriminatory and related outcomes are complex and shielded from scrutiny, including by contract and intellectual property laws. In some contexts, not even computer programmers may themselves be able to explain the way that their algorithmic systems function. This “black box” 99 effect makes it difficult for affected groups to overcome steep evidentiary burdens of proof typically required to prove discrimination through legal proceedings, assuming that court processes are even available in the first place. On the other hand, the companies responsible for creating and implementing emerging digital technologies face few if any legal requirements to prove that their systems comply with human rights principles and will not produce racially discriminatory outcomes. 45. International human rights law will by no means be a panacea for the problems identified in this report but it stands to play an important role in identifying and addressing the social harms of artificial intelligence, and ensuring accountability for these harms,100 as has been highlighted by the Special Rapporteur on the promotion and protection of the right to freedom of opinion and expression. 101 Ethical approaches to governing emerging digital technologies must be pursued in line with international human rights law, and States must ensure that these ethical approaches do not function as a substitute for development and enforcement of existing legally binding obligations. In thus section, the Special Rapporteur explains the concepts and doctrines of direct, indirect and structural racial discrimination under international human rights law and outlines the obligations they impose on States where emerging digital technologies are concerned. These obligations also have implications for non-State actors, such as technology corporations, which in many respects exert more control over these technologies than States do. This section also includes a nonexhaustive list of recommendations for concrete implementation of the norms and obligations presented. A. Scope of legally prohibited racial discrimination in the design and use of emerging digital technologies 46. The Special Rapporteur recalls that international human rights law is based on the premise that all persons, by virtue of their humanity, should enjoy all human rights without discrimination on any grounds. The prohibition on racial discrimination has achieved the status of a peremptory norm of international law 102 and as an obligation erga omnes. 103 Under international human rights law, States have further elaborated racial equality and non-discrimination obligations across a number of different treaty regimes; the principles of 98 99 100 101 102 103 14 Ibid. Frank Pasquale, The Black Box Society: the Secret Algorithms that Control Money and Information (Cambridge, Massachusetts, Harvard University Press, 2015). See https://datasociety.net/wp-content/uploads/2018/10/DataSociety_Governing_Artificial_ Intelligence_Upholding_Human_Rights.pdf. A/73/348, paras. 19–60. Human Rights Committee, general comment No. 29 (2001) on derogations from provisions of the Covenant during a state of emergency, paras. 8 and 13 (c). See also A/CN.4/727, para. 59. Barcelona Traction, Light and Power Company, Limited, Judgment, I.C.J. Reports 1970, p. 3, at p. 32, para. 33.