From Standards to Practical Applications
Strategies and standards are undoubtedly important when drawing the great lines that should inspire public policies. We have already referred to the developing strategies and their shortcomings. As for the standards, although to date there is no set of typologies that have been developed particularly for the region, around the world -and as a result of the efforts of national, public, private, academic and international entities- there are more than 80 bodies of principles, guides or technical standards referred to AI, as reported by international researchers.23 The recent announcement of UNESCO adds up to their perspectives, seeking to develop a set of principles that represents the United Nations vision on the subject, and for which it will be conducting a series of consultations that began with a meeting in São Paulo, in the which a community of experts from Latin America met this December.
Faced with the proliferation of standards, it is worth asking whether it is worthwhile to put additional effort into the development of more bodies of principles, or if those already developed, illuminated by a progressive interpretation of the international human rights framework are sufficient guidance. From our region, there may be a relevant opportunity to focus on a more practical approach to the subject, taking advantage of everything already debated and the common trends in the principles identified that speak of transparen-
19 https://www.mintic.gov.co/portal/604/articles-107147_recurso_1.pdf
20 https://www.gob.mx/mexicodigital/articulos/estrategia-de-inteligencia-artificial-mx-2018
21 https://36dc704c-0d61-4da0-87fa-917581cbce16.filesusr.com/ugd/7be025_9e91bfffeea647a0a663630 ea716aa8f.pdf
22 https://www.derechosdigitales.org/wp-content/uploads/ciberseguridad.pdf
23 Jobin, A., Ienca, M. & Vayena, E. The global landscape of AI ethics guidelines. Nat Mach Intell 1, 389–399 (2019) doi: 10.1038/s42256-019-0088-2, https://www.nature.com/articles/s42256-019-0088-2
cy, justice and equity, non-maleficence, responsibility and privacy, which are they repeat through different standards, although with divergence in their interpretation.24
From Derechos Digitales we propose an approach focused on the development of impact evaluation systems in the exercise of rights of the technologies of decision-making automation and AI, both from the perspective of public and private decision makers, as well as by agencies of international cooperation that in many cases finance the development of such technologies in the region.
As an organization, in this effort we have been linked to initiatives such as the B-Tech Project,25 led by the Office of the United Nations High Commissioner for Human Rights, which seeks to improve the quality of implementation of the United Nations Guiding Principles on Business and Human Rights regarding technology industries. Although the work is still in progress, it would benefit from a greater participation of representatives from Latin America, who can contribute to our visions of how the automated decision-making industries and their link with the states are impacting the exercise of rights in our region.
Far from immersing ourselves in a technosolutionist optimism utopia, dazzled by the capacity of these types of technologies, the authorities are obliged to open a democratic debate with the community regarding the aims of the technical solutions.
In this same sense, an initiative that deserves particular attention is the fAIr LAC alliance, led by the Inter-American Development Bank, which brings together representatives of the public, private and civil society sectors to ensure the responsible and widespread adoption of AI in Latin America and the Caribbean.26 The initiative emphasizes “taking advantage of the benefits of reliable artificial intelligence, which places the citizen at the center of decisions, while identifying their ethical and privacy challenges, and has mechanisms to mitigate them.” Although the focus on the ethical treatment of data and privacy considerations falls short for the reasons outlined above, it has enormous potential in creating concrete guidelines for evaluating the design, implementation and monitoring of projects to support the work of the governments in the implementation of automated decision-making systems and AI focused on inclusion and social justice.
Table of Contents
- Artificial Intelligence For What?
- Data Quality
- Ethics Is Not Enough
- Involved rights beyond privacy
- How Are Intelligence Strategies Being Developed In Our Region?
- From Standards to Practical Applications
- Artificial Intelligence from The South