Women of Uganda Network (WOUGNET) organized a Side Session on May 6th, 2022 during the Online Global Digital Development Forum (GDDF) 2022 to share perspectives on the interaction of data and social justice, informed by the data justice research conducted in Bolivia, Cameroon and Uganda by Internet Bolivia Foundation, AfroLeadership, and WOUGNET respectively with financial support from the Alan Turing Institute (ATI) and the International Center for Expertise in Montréal on Artificial Intelligence (CEIMIA). The conversation had an emphasis on privacy and social justice in terms of access to, visibility, and representation of data used in the development of artificial intelligence (AI)/machine learning (ML) systems, participation of the citizens in the decision-making, and management of databases. This session was submitted by Sandra Aceng, WOUGNET’s Program Manager; and facilitated by Peace Oliver Amuge, WOUGNET’s Executive Director.
The Definition of “Data Justice” in Different Regions
The session panelists included Eliana Quiroz, Research Director at Internet Bolivia Foundation; Louis Fendji (Ph.D.), Research Director on Innovative Technologies at AfroLeadership; and Isaac Amuku, the Research Specialist at the Women of Uganda Network who shared their definitions of data justice in Africa and Latin America.
Louis Fendji informed the audience that data justice refers to access to means to produce and control data. However, the challenge comes in how people are represented in data. He emphasized that data justice can have different meanings from one region to another. He cited that the diversity in sub-Saharan Africa is about change in data justice.
Isaac Amuku defined data justice as some of the forces (social, historical, economic, and political) that inform data collection, analysis, and use. These forces may influence how people are treated in terms of fairness, fair play, equity, peace, genuine respect for people, rightfulness, and lawfulness.
Eliana Quiroz shared that there is not an easy translation of data justice to Spanish, although people are living with several data injustices daily. This brings difficulty when making research on this topic. In order to build common ground with the people and be able to talk about data justice from that point, the use of examples was the best strategy.
The Interaction of Data and Social Justice including Data Justice Research Findings
The speakers gave their perspectives on the interaction of data and social justice including the data justice research findings, with emphasis on privacy and social justice in terms of access to, visibility, and representation in data used in the development of AI/ML systems, participation of the citizens in the decision making and management of databases.
The understanding of data justice can vary from region to region. Louis Fendji (Ph.D.), informed participants that data especially in the context of Cameroon is mainly collected by big tech companies due to a lack of national strategies by the government to collect data, produce and control data that is produced in the country including the lack of laws on data especially data protection and data privacy in Cameroon. In Cameroon, there is the existence of laws on electronic communications, and consumer usage of services however there is no law on privacy or data collection. This makes those with data have the power to do whatever they would like to do with existing data. There is therefore a power imbalance in the Cameroon context where developers should be aware of how data can be harmful to people that use their system. Additionally, people on the conflict side of Cameroon do not have the power to manage data.
In terms of the participation pillar, data can be harmful to people based on compliance because they do not have enough funds to support the understanding of the perspectives of marginalized communities on data because the process is expensive although developers are used to just collecting data and dropping it.
The platform for passport issuance by the government has some communities missing because their identities are missing. There is a need to make sure data is issued fairly to benefit all the communities.
Eliana Quiroz mentioned that the legal authority for example the communication authority accepted the reform of the unethical and legal ways in terms of internet service providers (ISPs). “We found power imbalance based on the knowledge of technology because if we want to say something as users, people are not involved in the ecosystem of governance, hence limited information and knowledge,” Quiroz added. This is because the government makes it hard for people to understand the debate as being technical and inaccessible. People are not included in the decision-making of data management. We have been putting pressure on the government to open data but there are gaps due to limited skills to collect, clean, and analyze data in terms of data presentation. There is violence against lesbian couples and the existing law cannot be used because the law favours violent crimes towards a woman by a man but not between two women.
Isaac Amuku spoke about the research by WOUGNET which aimed at broadening the understanding of data justice based on the six pillars (power, equity, access, participation, identity, and knowledge) among policymakers, impacted communities, and developers in urban, peri-urban, and rural areas of Uganda.
In the context of Uganda, there is the data protection and privacy act 2019, however, Non-Consensual Intimate Images (NCII) and Online Gender-Based Violence (OGBV) keep happening although no action has been taken because people in power do not act hence the power imbalance. Many people interviewed were moderately aware of data justice and the policymakers need to engage and know the perspectives of everyone to be integrated into the policy.
In terms of the equity pillar, it is usually the government collecting data and there is no clarity on where the data collected goes. For instance, the registration for a National Identity (ID) card is mandatory for everyone in Uganda to be able to access services although the data collection in Uganda is informed by culture and norms. This affects the LGBTIQ communities because they may not have certain services for being LGBTIQ and having missed data. The PWDs also have a lot of discrimination in terms of access to information and lack of representation when accessing the services.
In terms of the access pillar, there is no transparency in terms of surveillance done by the state because no one has access to this information collected from the citizens.
In terms of the identity pillar, there are so many identities of people collected who are not reflected. For instance, the refugees in Uganda do not know where their data go once collected.
Question and Answer Session
The question and answer session followed from the participants to the speakers. Participants asked:
In this era of datafication, I’m worried about Artificial Intelligence algorithms. Tech companies are using AI algorithms to increase productivity but these algorithms sometimes exclude the minorities and the marginalized communities. How can we influence tech giants and tech professionals to develop unbiased AI algorithms in the fight for social justice?
Eliana Quiroz: There are very few Apps in local languages and the elderly need support from younger relatives otherwise they do not benefit from information on the internet but they are also suffering from data collection. For example, Facebook collects data from users and non-users and yet elderly people do not receive any benefit from the data collected which is the same case as the government who are collecting data from citizens yet they do not benefit from it.
Article 3: Under the principles of Data Protection in Data Protection and Privacy act 2019
(1)A data collector, Processor, or Controller or any person who collects, processes, holds, or uses personal data shall; and (c) Collect, process, use or hold adequate relevant and not excessive or unnecessary personal data. Do you really see this being applied when it comes to data in the Ugandan context?
Isaac Amuku: The Ugandan law is okay and they have minimal standards on data collection but there is power misuse and participation in data collection such as in passport processes and National ID processes. There is a need for continued conversation on data protection and privacy to define data collection standards in Uganda.
A participant urged the need to bear in mind the universality aspect of human rights and have data justice inclusivity. In the Ugandan context, there is the misuse of data during data collection which is a violation of rights. He added that the means and ways to mitigate the effects of all processes of data collection are essential because the aspects of intersectionality need analysis in terms of data collection and planning. This is because a lot of data is left uncollected which affects the planning of the country.
A female participant added that most of the information collected by these companies can be easily used by third parties to acquire people’s personal data.
Louis Fendji (Ph.D.) said that data justice is about the representation of minorities and marginalized communities in the data. However, there is a need to ask ourselves why these communities are not included. Telecom companies are present in areas where they can get a return on their investments since they are profit makers. It is expensive to include these communities, especially for developers because developers sometimes lack funds to include marginalized communities in AI algorithms systems. However, developers need to be educated about the dangers of not including these communities and design national strategies to provide awareness to the policymakers about the inclusion of marginalized communities to be represented in data and have the power and ability to create the awareness.
YouTube Video: Introducing Data Justice
By Sandra Aceng – Manager Information Sharing and Networking