Rethinking Data Governance for Just Public Data Value Creation and Responsible Artificial Intelligence in Africa
The data-driven digital revolution presents an unprecedented opportunity for many sub-Saharan African (SSA) countries to harness digital resources that have potential to accelerate their socioeconomic development objectives. Consequently, data governance is a top policy priority for many African governments to maximise the benefits of data access and transfers, while addressing multilevel related risks and challenges.
Why do we need to rethink the current state of data governance in SSA?
A sound data governance framework goes beyond privacy and security considerations and acknowledges that collecting data alone has no value if it does not promote demand, usability, and impact, at scale. Robust data governance guides best practices for responsible, ethical data innovations particularly in the context of leveraging interdependent data-driven digital technologies such as artificial intelligence (AI) and machine learning (ML). A robust data governance framework is also an important component of enabling better quality and more granular data to achieve development goals, and ensure people’s digital rights are protected through policy tools and frameworks that ensure just public data value creation and responsible AI, that captures African values and contextual realities.
While many private companies can do more to share various forms of data for the common interest, and should be accountable for their role in extractive data practices at different points in the AI value chain , including the unprecedented wealth created by their unfettered data collection; the public sector is often a major data producer and collector and has an important role to play in improving public planning , service delivery, climate change mitigation efforts, and regional economic integration. Unlocking the economic and social value of public sector data through an enabling regulatory and policy environment can improve economic efficiency, may enhance democratic accountability, and encourage trust in the government.
In addition, high quality machine-readable data forms a powerful value chain, which can be used to create value (insights, intelligence, and products) with data dependent frontier technologies to create powerful analytics that can vastly improve innovation systems, sustainable digital development, public governance, and service delivery.
However, many data governance frameworks in SSA tend to focus too narrowly on the collection and production side of data under the assumption that whatever data produced will be used and data has inherent value. While many SSA governments acknowledge the importance of open data and data governance to ensure the design, collection, management, use, and re-use of data to foster robust trustworthy data ecosystems , these efforts fall short as they are often limited to privacy and security concerns which may inhibit scaling the positive and transformational benefits of data innovations for the public good.
As shown in the figure below, an effective data governance framework acknowledges that beyond mitigating the risks and harms associated with data, there are also considerations of creating public value from data. Reaping equitable benefits from data is highly dependent on acknowledging contextual realities to inform coordinated and transversal regulatory and policy frameworks that facilitate a conducive interoperable data ecosystem.
Figure 1: Framework to facilitate greater regulatory and policy coherence in a complex dynamic data ecosystem
Source: Authors own
There are multiple challenges that impact effective data governance in SSA data ecosystems, these include the following :
i. There is often limited funding to support state-led data curation to ensure data quality, active data demand, interoperability, and ongoing management of public data through its lifecycle,
ii. Sources of public administrative data are fragmented, scattered, and poorly organized;
iii. Multiple data collection initiatives implemented by non-state actors, that claim to be for the public interest are designed without co-creation and buy-in from state actors and funded in siloes resulting in data that is: collected independently, not acknowledged as part if the official national statistics system, or even used to inform official public policy making and thus inadequate to address cross-cutting developmental needs.
This is not a comprehensive list, the aim is to highlight that the value of public data for development in SSA is largely untapped, since realizing public data’s full value entails repeatedly reusing and repurposing data in responsible and creative ways to promote economic and social development. Therefore, a sound data governance framework requires that beyond privacy and security, institutions and stakeholders have the right incentives to produce, protect, use, re-use and share data along the data value chain, ultimately promoting just public data value creation.
What is just public data value creation and how can it support better data governance in SSA?
Just public data value creation denotes that data in itself has no value and existing power dynamics, exclusions and bias in data sets and data-driven ecosystems inhibit who benefits from public interest data-driven decisions. Just public data value creation emphasises a human-centred approach to funding, collecting, using, and sharing quality data for positive impact and data innovations that capture the multidimensional aspects of data as a digital public good (DPG), protect data subject’s privacy, support the social contract for data, and mitigates existing multidimensional inequities that arise and are exacerbated due to datafication of socioeconomic and democratic activity.
In addition to other inputs, my main contribution to the African Union Commission’s Data Policy Framework was to highlight the importance of enablers and governance needed to create public value from data, but there needs to be more collaboration and co-creation between policy makers and data practitioners to promote data governance frameworks that captures the components of the data value chain and how infrastructure, laws and regulations, policies, technical standards, and institutions impact just public data value creation at various public policy governance levels.
There needs to be more support, and funding for interdisciplinary localised research in African public policy and knowledge ecosystems that highlights the coordination amongst key stakeholder groups (i.e. private sector, civil society, academia and public authorities), political economy implications, and institutional structures necessary to promote just public data value creation that considers African realities, at local, national,regional, and global levels. Rigorous local research is crucial to encourage investments in critical infrastructure, understand the impact of datafication on socioeconomic activity, mitigate injustices that may be amplified by data, enable productive and socially valuable data innovations and uses, and ultimately capture the economically valuable characteristics of data as a factor of production to unleash wider social benefits and a fair data future.
