Link Search Menu Expand Document
Publications
  1. Data Quality

Data Quality

Automated decision-making and future AI developments in our region depend essentially on the existence and availability of a relevant cluster of data with which systems can be fed. The result of the automation processes can be radically different depending on the quality of the data provided.

In Latin America, this is a problem: ours is a continent in which discrimination by gender, race, income level, social status, educational level, age, sexual choice, among many others, is still sadly present in every

3 https://lpeblog.org/2019/11/25/the-second-wave-of-algorithmic-accountability/

daily decision, of the public and private world, and directly impacts the freedom and dignity of its inhabitants. The data collected in our countries is a faithful reflection of our political, economic, social and cultural systems. Starting the machine of the automation of decisions would imply conditioning the future to our past of discrimination and inequality. First, it is imperative to carry out a real and in-depth questioning exercise on how to generate the data that will allow us to meet the social justice objectives that should be drawn for these tools.

In this sense, it is interesting what an initiative like the A Plus Alliance4 proposes; led by Ciudadanía Inteligente and to which several organizations that work for digital rights in the region have joined. It is an effort that points out the need to critically examine the data from a gender perspective, to ensure that the implementation of algorithmic systems is an opportunity to correct the historical inequality faced by women. We need more initiatives of this kind, that look at the quality of the data as a prerequisite to any development of any tool that honestly pursues social justice objectives, to avoid more mirages of progress in the region.

Table of Contents