Disclosure: The views and opinions expressed herein belong solely to the authors and do not represent the views and opinions of crypto.news editorials.
Today, AI is ubiquitous, impacting everything from increasing productivity in the workplace to how we solve emotional and personal problems. Innovation in this sense has its benefits, but it lacks the ability to have a tangible impact on the most disadvantaged parts of the world.
summary
Centralized AI will fail the Global South, reinforcing biases, eroding data sovereignty, and creating opaque and unaccountable systems that undermine the United Nations Sustainable Development Goals. Decentralized AI powered by federated learning and blockchain provides a comprehensive, secure, and transparent alternative that enables real-world deployment across local data control, responsible governance, climate change response, healthcare, payments, and conservation. The path forward requires moving from corporate AI to open, decentralized infrastructures that embed inclusivity, sovereignty, and accountability into that architecture so that AI can ethically contribute to global development.
The United Nations Development Program has worked tirelessly to pursue its 17 Sustainable Development Goals to end poverty, advance climate action and support equitable growth by 2030. Given existing global use cases and applications, it would be natural to see AI as key to fostering inclusivity and global development. However, the current centralized architecture of AI is plagued by issues such as data privacy concerns, high costs, and limited access, severely impairing AI’s ability to last forever.
This centralized nature of AI will only reinforce existing power imbalances, and its inherent biases, stripping of data sovereignty from communities, and lack of transparency will prevent AI from realizing its potential for good in the Global South. For AI to be effectively applied as a tool to drive global development, there needs to be a shift from a centralized enterprise architecture to one based on inclusivity, sovereignty, and accountability. Decentralized AI is the solution.
The paradox of centralization
While AI is already being applied to solve challenges from climate change to healthcare, the reality is that its development is largely dominated by a handful of tech giants, with systems that are substantially and unethically inappropriate for the unique context of the United Nations’ 17 SDGs. But the crisis is not a technology problem; it is a governance problem. Traditional AI development models create three different obstacles to true development impact.
Most of the centralized models are trained on data from a small number of developed regions, excluding the entire Global South. Research shows that these models, when deployed in a variety of contexts, such as diagnosing diseases or predicting financial risks, become fundamentally inadequate for their intended purpose. A lack of appropriate training can lead to systematic misidentification, denial of critical services and reinforcing socio-economic disparities, posing a threat to SDG 10, which aims to promote social, economic and political inclusion for all.
These systems also require highly sensitive local data, from patient records to financial and criminal records, to be aggregated onto remote corporate servers that are centralized and therefore susceptible to hacking. Data extraction practices deprive governments and institutions of their right to data sovereignty, threaten SDG 16, which asserts the right to peace, justice and strong institutions, and jeopardize the security of data aggregated from local servers. This practice has also led to the proliferation of sovereign AI technologies that are emerging in countries such as Singapore and Malaysia in an arms race to maintain data sovereignty.
But the most important consideration is who will be held responsible if opaque and poorly understood AI makes serious errors or generalizations about policies that could impact millions of lives. The “black box” nature of centralized AI systems, including their ownership and mechanisms, makes it difficult to make auditing decisions such as allocating aid, modeling risk, and assigning accountability, making them ethically unacceptable for high-stakes development efforts. This lack of transparency could undermine all 17 SDGs.
The only way to reconcile the power of AI with the ethical requirements of international development is through a fundamental shift from centralized corporate AI training to mechanisms based on the principles of inclusion, sovereignty, and accountability.
Decentralized AI: A two-pronged solution
Decentralized AI based on federated learning and blockchain technology is emerging as a solution to this challenge. The SDG Blockchain Accelerator Program, strategically led by UNDP and supported by partners such as the Blockchain for Good Alliance, Stellar, FLock.io, and EMURGO Labs, further exemplifies this by pioneering decentralized AI initiatives that empower rather than hinder communities in the Global South.
Federated learning works by training a shared model across multiple distributed devices while preserving local data. Projects in Latin America and the Caribbean will use this technology to collaboratively train predictive AI to accurately predict climate risks while keeping regional financial and demographic data securely on local servers. This infrastructure will support efficient and fair payments to climate-vulnerable farmers and women-led businesses, achieving both SDG 13 (climate action) and SDG 5 (gender equality).
Operations supported by federated learning are complemented by blockchain technology, which replaces a single corporate intermediary with an immutable and transparent governance system. This creates an infrastructure built for collaboration and restores necessary accountability mechanisms. In Liberia, smart contracts and decentralized AI are being deployed to facilitate the transparent distribution of payments and aid, while in Kenya, decentralized AI is eradicating payment gaps for local businesses and increasing economic growth and trust in public institutions outlined in SDG 8 (Decent Work and Economic Growth) and SDG 10 (Reducing Inequalities).
Additional applications of decentralized technology to support the SDGs include the development of blockchain-based NFTs by the University of Cambridge and UNDP Rwanda to help conserve mountain gorillas, and the security of hospital records in Africa, giving patients the autonomy to grant or revoke access to their patient records to achieve SDG 15 (Life on Land) and SDG 3 (Health and Well-Being).
Architectural liability requirements
AI holds immense potential as a technology, but its fundamental challenge is one of governance. While centralized proprietary models fundamentally undermine the principles of inclusivity, sovereignty and accountability that the UNDP SDGs embody, current work shows that viable, ethical and scalable alternatives exist.
The challenge now for the global development community is to prioritize funding for deploying open, decentralized AI infrastructure over corporate tools that inhibit development. It is time to change our mindset from that of passive consumers to stewards of intelligence that promote a sustainable future for the most disadvantaged parts of the planet.
