“As feminist researchers with an interest in technology, the question we should ask is, how does the research contribute to liberation and transformation of technology to be used in its full capacity by women, gender diverse, and vulnerable groups on basis of race, sexuality, caste, ethnicity etc.?’
We conduct research for evidence in policymaking; to determine how the solution we are developing will work and to assess the impact of our solution. There are several ways to skin this cat called research – whether through a numbers game (quantitative) or through an experiential approach (qualitative). One can also start with a theory that they seek to test and prove or disprove (inductive) or start with questions and let the research guide you developing theories based on the patterns observed (deductive). There are of course debates on what makes good research – the numbers or the words- but this is not the focus for now.
In the first article of this series – I emphasised the need for gendered concerns on privacy and data protection to be considered in policy development. In this second article, I dive into why a feminist methodology is useful to think about AI, privacy, and data protection. I walk you through reflections on my feminist conceptual framing for my project on AI, the right to privacy and data protection in South Africa[i]. Does a feminist lens bring a different understanding of issues?
Why a feminist approach?
The replication of existing inequalities, development of new social injustices and unequal power dynamics impact the difference in experiences of decisions made by algorithms from AI and machine learning systems. A feminist lens helps to unpack the spectrum of issues and opportunities from this technology in a context of gender inequality. A feminist approach helps to answer the questions - why gender is even important in the conversation of privacy and data protection
In taking on a feminist perspective as seen in the first series – I was able to point out ‘uneasy access’ women and gender diverse people experience in the online space as an extension of patriarchal perceptions and control of their visibility. The emerging issues – include dataveillance – from private entities, business and society; bias and discrimination; violations of human dignity; the removal of agency and loss of control in interacting with technology and the power dynamics reflective of digital inequalities shaped by offline realities.
I was able to point out ‘uneasy access’ women and gender diverse people experience in the online space as an extension of patriarchal perceptions and control of their visibility.
In this world we exist in, that prioritises knowledge and where collaboration is best to get a point across, a movement is important to push these questions. The A+ Alliance for Inclusive algorithms provided that necessary space to ask why a feminist approach within its pilot of the Feminist AI Research Network. The focus of the network is to understand what can be done to ensure innovations that correct real-life bias and barriers preventing the participation of women and gender diverse people from fully participating and enjoying their rights[ii]. The feminist research questions asked in the alliance are:
A feminist conceptual framework - a cauldron of justice, feminist principles, and intersectionality?
Continuously responding to why a feminist lens in my research approach has refined my response to "it’s the only way" because a critical approach is needed to understand the impact of new technologies on women and gender diverse people in the context of existing social injustices. The question next is how then does one apply feminist thinking to understanding the areas of concern in AI, privacy and data protection? The frameworks we choose in our work influence the process and analysis of research.
In retrospect, there isn’t one concept sufficient enough to capture the diversity of issues that were being explored in this study. In conversations with feminist researchers – what was clear is that we may need to take from different places to build a conceptual framework that would work in a particular context and navigate this limitation[iii].
In retrospect, there isn’t one concept sufficient enough to capture the diversity of issues that were being explored in this study.
I drew my conceptual lens from data feminsim, intersectionality, data justice and feminist principles of the internet. The conceptual lens allows one to ask questions of who is being represented and by whom; whose interests are being centered; why this discussion is important and how it is taking place, which allows for criticism of power and how data itself can be used to ensure justice in society[iv].
A feminist conceptual lens
7 guiding principles of data feminism
The principle of privacy and data that supports the right to privacy and full control over personal data and information online at all levels
An understanding of social interactions with technology that consider multiple inequalities and locating technology in the context of systematic oppressions including racism, sexism, colonialism, classism, and patriarchy
A way to center marginalised groups in datafied societies to recognise opportunities and respond to harms emerging from use of data in society that may have an adverse impact on groups in society.
A feminist guiding tool with consent, accountability, respect, privay, safety and reciprocity the main ethical concerns of this study
Where does this knowledge come from? – embodying multiple points of knowledge
Context underpins feminist research and I have been focused on gender inequality as this is the reality of many South African women and gender diverse people at the intersection of race, wealth, sexuality, and class etc. The conceptual lens guided my methodology to break binaries of gender, prioritise multiple points of knowledge and centre the experiences of marginalised groups rather than the technology itself. A mapping study of the AI policy landscape indicates that the conversation about gender was taking place around participation in the development of systems and upskilling. Qualitative interviews and a quantitative purposeful survey unpacked more of the societal issues.
Ten in-depth interviews were done with key informants who understood the context of the law and considerations of privacy and data protection in the context of AI. The highlight of these interviews was the challenge of current policy and regulation in engaging with gender issues and being responsive to contextual realities. The survey with 25 gender and sexual justice activists gauged the perceptions of AI, privacy and data protection issues from activists who work with communities that are affected by these technologies. This survey helped to fill the knowledge gap of the lack of documentation of AI impact in South Africa by building up perceived concerns of data harms based on scenarios of AI harms mapped in the research.
Conclusion - Feminist deductive work
In conclusion, feminist methodology provides with it intentionality to critically look at technology in context and as part of the solution. There are several other framings that one can use to engage with issues of social justice; feminism allows for me to centre women and gender diverse people, situate the knowledge developed, ask critical questions around power and reflect on my own journey through the process. In the final blog series, I will be writing "A call to action from an African feminist" – after all, we need to make a feminist internet.
Interested in following this project – please fill in this quick survey
[i] South Africa exists in the context of a triple threat: inequality, poverty, and unemployment. As a result, it requires a more critical approach to the 4thIR. The study assess the adequacy of privacy and data protection in South Africa in order to develop a gender responsive guide on safeguarding against AI harms and making use of technologies to benefit society.
[iii] Thank you to Anri van Der Spuy, Maya Ganesh and Tigist Hussen who helped me navigate conceptual framework outside of an academic space but with some form of academic rigour.
[iv] C D'Ignazio and LF Klein Data feminism (2020).