Afro-Feminist AI Governance: Challenges and Insights from the 2022 Internet Governance Forum
At the 2022 Internet Governance Forum, Pollicy had a productive and thought-provoking discussion on the emerging field of Afro-Feminist AI Governance. As AI technologies become increasingly embedded in everyday life, and especially in the Global South, there is a need to ensure that these technologies are designed, deployed and managed in a way that is respectful of the rights, experiences and perspectives of African women and girls. In this blog post, we reflect on the key challenges and lessons from our session at the Internet Governance Forum related to Afro-Feminist AI Governance.
At the 2022 Internet Governance Forum, an important discussion was held on the need for Afro-Feminist AI Governance: Challenges and Lessons. The topic is not only relevant to global AI governance, but it is also vital in terms of addressing the distinct needs of African women in AI, who have been largely excluded from the development of AI systems and decision-making processes.
To start with, Afro-feminist AI governance is simply a framework for regulating and guiding the development and use of artificial intelligence (AI) that prioritizes the experiences and perspectives of Black women. This approach recognizes the unique challenges that black women face in relation to AI and seeks to address them in a way that is inclusive, intersectional, and empowering.
One of the key challenges of Afro-feminist AI governance presented during the session, is the lack of representation and inclusion of black women in the field of AI as cited by Pollicy in the Afro-feminist data futures report. The tech industry has long been dominated by white men, and this is especially true in the field of AI. Black women are often excluded from leadership roles and decision-making processes in the development and deployment of AI technologies, leading to a lack of diversity and inclusion in the field.
This lack of representation has real-world consequences for black women. AI technologies, like many other forms of technology, can reinforce and amplify existing biases and inequalities. Without the input and perspective of Black women, AI systems are more likely to perpetuate harmful stereotypes and discrimination against black women.
For example, AI-powered facial recognition technologies have been shown to be less accurate at identifying black women than white men, leading to potential consequences like wrongful arrests and other forms of discrimination. Without Afro-feminist AI governance, these biases and inequalities are likely to continue and even worsen.
Another challenge of Afro-feminist AI governance mentioned by Amber Sinha during the session, is the need to balance the potential benefits of AI with the potential risks. AI technologies have the potential to greatly improve many aspects of our lives, from healthcare and education to transportation and communication. But they also have the potential to cause harm, whether through the amplification of existing biases and inequalities or through more direct forms of harm like loss of privacy or job displacement. As these systems are largely developed in a western context and then applied in Africa, they often do not take into account contextual factors and have serious discriminatory and exclusionary impacts.
Afro-feminist AI governance seeks to reframe this discussion by prioritizing the experiences and perspectives of black women. According to Zulfa Bobina, This means taking into account the unique challenges that black women face in relation to AI, and ensuring that the development and deployment of AI technologies takes those challenges into account at each stage of development and implementation.
One way to do this is by incorporating a diverse range of voices and perspectives into the development and deployment of AI technologies. This can include not only Black women, but also other marginalized groups who are often excluded from the tech industry, such as people of colour, people with disabilities, and sexual minorities.
Another way to address the challenges of Afro-feminist AI governance is by promoting transparency and accountability in the development and deployment of AI technologies. This means ensuring that AI systems are designed and implemented in a way that is open and transparent, and that there are clear mechanisms in place for holding AI developers and users accountable for any harm that their technologies may cause.
For example, emerging experiments around social transparency, Human Centred XAI, regulatory proposals on risk assessment in legislations such as EU’s AI Act and appropriate metaphors of data which go beyond the hackneyed comparisons to oil such as data as bodies and decisional autonomy need to be looked at.
Overall, Afro-feminist AI governance is a crucial approach for addressing the unique challenges and perspectives of black women in relation to AI. By promoting representation, inclusion, and transparency, Afro-feminist AI governance can help to ensure that AI technologies are developed and used in a way that is fair, equitable, and empowering for black women and other marginalized groups.
Written by Ibrahim Waiswa Batambuze for Pollicy.
