Mark Latonero: Data-Driven Tensions in Human Rights

During the GovLab’s most recent Ideas Lunch, Mark Latonero, Fellow at the Data & Society Research Institute, spoke about the numerous tensions that result from data collection in human rights issues, emphasizing both its potential for good and the dilemmas it provokes.

Picture1

In the past decade, a growing group of innovators began finding solutions in the development and humanitarian field by harnessing advances in data-driven technologies, but often faced skepticism from human rights advocates regarding the capacity of new tools to meaningfully affect conditions on the ground. Nonetheless, on September 25, 2012, President Obama gave a speech declaring that the government would harness innovations in technology to combat human trafficking. Since then, data use has come to the forefront of human trafficking efforts as a means of both identifying and countering human rights violations. While data-driven approaches to addressing human rights issues have the potential for significant good, Latonero explained that there are tensions and tradeoffs to be carefully analyzed in order to ensure that technology is being used appropriately.

Latonero used his last six years of work in investigating cases of online human trafficking as an example of how data driven applications can be used in a human rights context. As individuals have moved to online communication, so too has human trafficking. As a digital ecosystem easily connecting supply and demand within the sex industry emerged, so too have some cases of commercial exploitation of children hidden within online venues such as “classified” sites, like Backpage. Counter-trafficking efforts have primarily involved manually scanning such sites for potentially illicit cases of commercial sexual exploitation of minors, but the process is tedious and inefficient. Latonero and his team used database integration, data markup, and visualizations to help identify high-risk posts in a fraction of the usual time. On a wide-scale, this sort of data analysis could not only improve response from authorities and mitigate human rights violations, but also save huge amounts of time and resources.

Despite potential benefits, Latonero warned that there are risks and challenges that need to be considered as data collection and data-driven approaches are implemented in an attempt to bolster human rights on a large scale. He identified four key tensions that have yet to be carefully analyzed by researchers and policymakers:

Monitoring and privacy

Observing human rights issues through data can lead to risks associated with increased surveillance. Government surveillance – in the wake of Edward Snowden’s NSA leaks – is a particularly fraught public issue. “Monitoring,” rights abuses carries positive connotations like protection and security, whereas “surveillance” carries negative connotations like intrusion, control or punishment. A priority in using data-driven technologies in human rights efforts, then, is to ensure that public policy and decision-makers harness such technologies in responsible ways.

Temporal tensions

Data persists in time, and information that was once essential to stemming a crisis can later marginalize or otherwise harm an individual. In many high-risk situations, the public makes a tradeoff of certain values, like privacy, for safety. After the crisis is resolved, though, the public may suffer. In the case of Ebola, for example, efforts to contain the disease included distributing information on potential carriers. Once the outbreak ended, though, some of those high-risk individuals were publicly stigmatized. How, then, should values fluctuate in confronting an immediate problem, if at all?

Picture2

Risks, harms, and benefits

Data-driven applications have not been implemented on a wide scale for an extended period of time. The potential for misuse is great, and if implemented by repressive regimes could be a tool for significant harm. What are the long-term risks of developing surveillance techniques? What are the benefits? Should data be aggregated or can it be used to identify individuals?

Context, power, and implementation

There is currently no framework that governs data use as implemented in human rights issues.  Because rights issues are highly contextual, it is difficult to create universal standards for implementation and policy making. How can we take into account local context and power relationships in developing a framework for action? How can data-driven technologies be leveraged to empower as well as protect vulnerable populations?

To help address challenges in data-driven human rights efforts, Latonero called for a multi-stakeholder approach to the analysis of these tensions. Additionally, he highlighted the need for “translators” between the multiple stakeholders involved. The involvement of data developers in a conventionally government-controlled domain creates the potential for miscommunication, as policymakers may not understand the technology being used and data scientists may not understand the implications of their developments. In order to best use these technologies, Latonero explained that the multiple actors involved (policymakers, government officials, NGOs, human rights activists, data scientists, engineers, etc.) must form a network bridged by actors who can translate the needs between them.

Data-driven applications are becoming more common-place in countering human rights violations, with numerous benefits.  From confronting sex trafficking to containing pandemics, data can assist in a variety of important initiatives.  But the costs, too, must be considered.  As Latonero looked toward the future of human rights efforts with optimism, he made clear that society must evaluate the tradeoffs they are willing to make in order to protect vulnerable populations from harm.

 

For more information see Data & Society’s Data and Human Rights Initiative and primer

http://www.datasociety.net/initiatives/data-human-rights/

http://www.datasociety.net/pubs/dhr/Data-HumanRights-primer2015.pdf

Mark Latonero can be found at @latonero and latonero.com

No comments yet.

Leave a Reply