A/HRC/29/37 good for conflict prediction and prevention, it has been less effective for analysis and actionable transmission.29 63. Big data mining and remote sensing work for the prevention of human rights violations also raises methodological and ethical concerns. For example, vulnerable populations could be put at risk by the remote documentation, and thus identification, of their locations and situations.30 Moreover, potential inaccuracies in the statistical analysis of human rights data arise from selection bias, duplication and constraints on data capture.31 4. Towards digital due diligence 64. The application of surveillance to prevent violations of human rights may be so effective as to imply that States with the capacity to take advantage of them have a responsibility to do so. Cameras have been used in police vehicles and interrogation rooms and consideration might be given to other contexts in which such surveillance could have a preventive effect (for example, prisons), subject to the limitations imposed by other rights, such as the right to privacy. 65. Other affordances of ICTs can be harnessed by States to fulfil their responsibilities concerning prevention or precaution. For example, there have been instances of States using text messages or calls to warn civilian populations before launching air raids. The recording devices on certain advanced weaponry offer the potential for greater oversight, but that will require more transparency. 66. In the digital space, however, the responsibility of due diligence extends beyond States. Human rights monitoring organizations — both intergovernmental and nongovernmental — need to give thought to the consequences of their correspondence or use of information. Traditional understandings of “informed consent” may need to be revisited. D. Monitoring and fact-finding 67. As noted above, the particular nature of violations of relevance to this mandate place a premium on fact-finding. Human rights organizations have developed rigorous factfinding methodologies, not least to protect the credibility of their evidence, and thus their reputations. ICTs and the user-generated content they facilitate have broadened and democratized the process of fact-finding by empowering both spontaneous and solicited “civilian witnesses.” The most challenging dimension of this evolution is balancing democratization with a continued, perhaps heightened, requirement for authority and thus for verification of digital evidence. 1. Civilian witnesses and video evidence 68. The advantages of video evidence have been appreciated by campaigners for several decades, at least since the Rodney King incident in the early 1990s. The re-purposing of private closed-circuit television footage for public investigations has become 29 30 31 12 Sheldon Himelfarb, “Can big data stop wars before they happen?” (United States Institute of Peace, 25 April 2014), www.usip.org/publications/can-big-data-stop-wars-they-happen. See, for example, Harvard Humanitarian Initiative, “The Signal Program on Human Security and Technology” (2013), http://hhi.harvard.edu/programs-and-research/crisis-mapping-and-earlywarning/signal-program. See the work of the Human Rights Data Analysis Group, https://hrdag.org/coreconcepts/.