Nubian Rights Forum & 2 others v Attorney General & 6 others; Child Welfare Society & 9 others (Interested Parties) [2020] eKLR 28. Mr. Anand stated that upon review of the Petition by the 1st Petitioner, the 1st and 2nd Respondents’ replying affidavits, and the relevant accompanying documentation, he had made several conclusions on NIIMS. The first was that NIIMS is functionally and architecturally similar to the Indian Aadhaar system. His second conclusion, based on the first, was that it would result in the same outcome, namely an endeavour which would pose a massive risk to personal security and privacy of the Kenyan residents with no demonstrable benefits. Lastly, that it would also create national security risks to Kenya, which would be impossible to mitigate. 29. Mr. Anand stated that a foundational assumption underlying the logic of the NIIMS system is that the use of biometrics assures that there is no duplication in the registration of individuals. Further, that the assumption behind biometric identifiers is that there exists a part of the human body that is unique to an individual, and we can use a digital representation of that part to uniquely identify the individual from across the entire population. It was his evidence, however, that biometrics are fallible and cannot be relied upon as unique identifiers for the purposes of deduplication. His averments, as we understand them, are that firstly, to claim that a particular biometric parameter is unique across all time is mathematically infeasible. This is for the reason that one would need to have data for past and future populations to be able to reach this conclusion, which is not possible for future populations. 30. Mr. Anand stressed, secondly, that biometrics are affected by ageing and time, and that every single cell and part of the human body ages and changes, and events in life such as disease or constant manual work can accelerate it. It was his averment therefore that the assumption of uniqueness with respect to biometrics is true only within a time range. He nevertheless noted that biometrics are stable after the age of 15 years, but “wildly variable before that.” 31. It was his averment, thirdly, that the identification of a person by a human being is a different process from that of identification by a machine or algorithm, and that human beings can identify others accurately because of the ‘evolutionary pressure to reproduce.’ It was his view therefore, that while a human being cannot fail to identify the gender of another human being, an algorithm uses different means to identify people through computation and has to use different conceptual constructs. 32. Mr. Anand explained in technical detail the concepts that are used in algorithms, including the use of points in fingerprints matching, to illustrate that a perfect overlap of points and matching of fingerprints is impossible because of extraneous factors such as sweat, tears, physical position and other mechanical factors. It was his averment therefore that thresholds are specified to identify a person through such matching of points. Mr. Anand’s contention was that there is an inherent conflict with the said thresholds for the reason that if a very high threshold is set, such as 99% matching of the points, it will enable a person to enroll twice at different points of his life, because ageing changes biometrics. On the other hand, that if the threshold is low, then some enrolments will be identified as duplicates and genuine persons would not be registered. 33. Mr. Anand’s testimony was that this is a problem that arises from the nature of biometrics and cannot be worked around or fixed. Our understanding of his argument, which was made in relation to the Aadhaar system in India, is that a person who had not previously enrolled or registered in the system could be identified as having been registered and rejected. Conversely, a person who was already enrolled could, after some lapse of time, be identified as not having been enrolled, and allowed to register a second time because of ageing biometrics. His assertions on this point were based on an article he had published on the internet on the Aadhaar project, the relevant website address of which was given, which made reference to a paper published in the Economic and Political Weekly by the Centre for Internet and Society (CIS). It was his contention, on the basis of this paper, that enrolment rejection rates will accelerate as the size of the database increases. 34. Mr. Anand gave an example of the errors of duplication noted in the Aadhaar system which is operated by the Unique Identification Authority of India (UIDAI). He observed that UIDAI used OT Morpho’s algorithms, the same algorithms which, to his knowledge, are also being used in the NIIMS in Kenya. He placed reliance on an http://www.kenyalaw.org - Page 7/174