Resource Center / Reports and Surveys / Building Ethics into Privacy Frameworks for Big Data and AI
Building Ethics into Privacy Frameworks for Big Data and AI
This report provides an overview of how organizations can operationalize data ethics.
Published: October 2018
Big data, new technologies, and new analytical approaches, if applied responsibly, have tremendous potential to be used for the public good. Big data’s greatest value for global development lies in harnessing the power of real-time and predictive analytics for smarter decision making, anticipatory approaches to managing risk, and new ways to measure social impact.
At the same time, big data amplifies risks to privacy, fairness,equality, and due process. Large-scale data collection can expose characteristics and behaviors of individuals, lead to biased decision making based on unrepresentative or inaccurate data samples, and lack transparency, preventing individuals from exercising due process rights. Organizations have to weigh the benefits of processing data against associated risks, but they o en lack a recognized systematic framework to deploy. Also, understanding the complexities of big data analytics requires the engagement of professionals who are versed in an interdisciplinary body of knowledge covering law, technology, ethics, engineering, risk management, and design.
The United Nations Global Pulse and the International Association of Privacy Professionals explored these issues and discussed their many aspects at a jointly hosted event: “Building a Strong Data Privacy and Ethics Program: From Theory to Practice,” held in May 2017 in New York (hereafter referred to as the “UN GP/IAPP event”).
The event brought together humanitarian and development organizations, including the World Food Programme (WFP), the United Nations Development Programme (UNDP), the United Nations High Commissioner of Refugees (UNHCR), the United Nations Children’s Fund (UNICEF), and the International Committee of the Red Cross (ICRC); NGOs; representatives of private sector corporations such as MasterCard, TrustArc, Nielsen, and IBM; as well as privacy regulators from across the world. The presence of stakeholders from different sectors presented an opportunity to view big data analytics not only from the perspective of routine uses of big data for marketing and other business-related purposes, but also in the context of humanitarian and development causes, such as using data to track the emergence and progress of pandemics, to prioritize resource allocation for sustainable development, or to target disaster relief to the most vulnerable populations.
This report provides an overview of how organizations can operationalize data ethics, drawing on the discussions at the UN GP/IAPP event as well as on additional research about data ethics and privacy best practices in a world of big data analytics.
It demonstrates how organizations deploying data analytics and artificial intelligence (aka AI) can reflect ethics considerations in their decision making, borrowing tried and true operational tactics from the field of information privacy. Such steps may include, for example: building a multi-disciplinary team across departments to practice ethics “on the ground”; conducting ethics assessments for new big data projects to consider their personal and societal impacts, while consulting with external and internal ethics working groups; and building programs that are scalable and flexible, which depend on factors such as the societal context of a big data project and the organizational structure of the entity performing it.
Approved