Can AI help us achieve the SDGs?

As momentum on the SDGs stalls, AI’s promise of exponential growth could offer much-needed rapid acceleration across the 2030 Agenda. To harness AI effectively, we must ensure it serves those most in need, and that all countries – not just those in the Global North – can tap into its development benefits

Data and monitoringGlobal

Satellite image of a Mediterranean hurricane approaching Greece. The European Space Agency is using AI in its Destination Earth project, which models the planet's climate system. ©ESA

The whirlwind of artificial intelligence (AI) continues to sweep across headlines, promising revolutionary advancements in everything from self-driving cars to personalized medicine. But amid the hype lies a crucial question: can AI become a powerful tool for tackling some of the world’s most pressing development challenges?

The answer is yes – potentially.

Used well, AI could significantly accelerate progress toward the Sustainable Development Goals (SDGs). These 17 ambitious goals, set by the United Nations, range from eradicating poverty to combating climate change, and aim to achieve a more sustainable future for all. 

Let’s delve deeper into how AI might be harnessed to make a difference.

Combating climate change (SDG 13)

The fight against climate change could receive a critical boost from AI. Climate modeling is one obvious area where AI is already making rapid advancements possible. Complex climate modeling initiatives like Destination Earth (DestinE), led by the European Space Agency, are providing a deeper understanding of our planetary climate system. AI-powered tools like FireAId, developed by the World Economic Forum, are helping predict wildfires and enabling more effective responses. By analyzing vast amounts of data and identifying trends, AI could empower researchers and policymakers to develop effective solutions in both climate mitigation and adaptation.

Transforming healthcare and education (SDGs 3 and 4)

The field of education is witnessing a surge in the use of AI-powered tools for personalized learning. These intelligent systems can act as virtual tutors, assisting educators and pupils alike by tailoring learning experiences to individual student needs. This not only caters to different learning styles but could also lead to improved student engagement and outcomes. This approach to learning could become especially important in resource-restrained areas in lower-income countries and help halt the growing “postcode lottery” of education on a local, national, and international scale.

Similarly, in healthcare, AI is revolutionizing diagnostics through image recognition for diseases like cancer. AI can manage vast amounts of patient data and enable faster and more accurate diagnoses. Additionally, AI-powered “wearables” and personalized medical devices could potentially lead to better health management and improved patient wellbeing.

Eradicating poverty (SDG 1)

AI is offering innovative ways to measure poverty more effectively. New technologies are capable of analyzing huge amounts of data from diverse sources – satellites, mobile phones, and digital finance records, for example – to identify poverty pockets with greater precision.

This, together with AI systems to predict the socio-economic status of potentially vulnerable populations, allows for better targeting of social protection programs to reach those most in need of support.

A joined-up approach: unveiling connections and monitoring progress

AI’s potential extends beyond tackling individual goals. It can also help us identify crucial links between seemingly disparate issues and, in turn, inform more effective policymaking.

For example, machine learning techniques have helped make it possible to map household wealth estimates onto daily temperature variations across more than 130 low and middle-income countries. This has established a significant relationship between the two variables: that higher temperature variability results in greater poverty

This opens up a whole area of policymaking around climate risk insurance for those most vulnerable to temperature variability and other climatic shocks – something that may have been overlooked if AI hadn’t enabled the connection between these variables to be made. It also places greater urgency on addressing climate change at an international level as a way of tackling global poverty.

As well as accelerating progress toward the goals, AI can also play a vital role in monitoring progress, helping inform policy decisions much closer to real time. For example, techniques like remote sensing, which involves analyzing data collected by satellites and other airborne platforms, can be coupled with AI to pinpoint deforestation, gather data on buildings, and assess damage after disasters. This allows for a more comprehensive and timely understanding of progress toward environmental goals.

AI is also supporting advanced statistical modeling techniques like “small area estimation.” This allows the generation of granular estimates for measures like poverty at a much lower cost and with greater frequency than traditional methods. It means policymakers can track progress toward specific goals with greater precision and make data-driven decisions.

No silver bullet

While AI presents immense potential, it’s important to acknowledge the challenges associated with its implementation. There are several issues which could slow – and in the worst cases negate – progress toward achieving the SDGs.

There are general risks associated with modeling that may be amplified by the use of AI technologies. For example, a concentration of funding and infrastructure in wealthier countries may mean models (such as those predicting climate) and their outputs are skewed toward the biases and aims of institutions in the Global North.

What’s more, regardless of the balance of power in terms of technology and research, there remains the need to understand bias in pre-trained models – as well as the need for more “real world” data to train models in the first place.

We also need to consider the huge levels of energy consumption associated with large AI models and modern technology stacks (such as blockchain). Unless mitigated in some way, this will lend an irony to the use of AI in tackling climate change.

Likewise, for education purposes, AI relies on technology which may be unavailable in lower-income countries. In this way, it may act to widen disparities, introducing a two-tier system in which only pupils in richer countries benefit from personalized learning underpinned by AI.

Ethical considerations regarding data privacy and security are also paramount, especially when dealing with sensitive personal information – in the field of healthcare, for example. Policies and regulations need to be established to ensure data is collected, stored, and used responsibly. Importantly, these regulations need to be consistent in scope across country boundaries.

Finally, ensuring transparency in decision-making becomes arguably more crucial in a world where policies are potentially underpinned by AI-assisted evidence. When it comes to targeting social protection, for example, the use of AI (to model poverty levels, perhaps) could be particularly problematic. We may begin to see a “black box” effect where households who don’t get selected for social protection support are not able to find out why.

The road ahead: navigate wisely

The role of policymakers and funders is critical in harnessing the power of AI for the world’s good. They need to direct investments toward areas with the most potential impact. At the same time, they must use their power and influence to address issues like “northern bias” and “algorithmic colonialism” to ensure that progress is meaningful. Civil society also has a role to play in advocating for international collaboration, responsible AI development, and equitable access to technology.

In conclusion, AI holds immense potential for accelerating progress toward a more sustainable future. By addressing the challenges, investing wisely, and ensuring ethical considerations are at the forefront of any implementation, we can harness the power of AI to bridge huge development gaps and create a more sustainable future.

Share post:

Related articles

Good intentions aren’t enough: the need for effective SDGs policy implementation

SDG-related initiatives are increasingly under attack from the more right-wing, populist parts of the political spectrum, who portray the Goals as misguided, wasteful, vain, and coercive. How can cities and local governments enhance their policymaking and implementation, and – crucially – secure buy-in to ensure initiatives achieve their intended outcomes?

Stefano Marta

Is clean technology transfer an empty promise?

Technology transfer is fundamental in developing countries’ aspirations to decarbonize, yet the flow of green tech from developed nations is far below what’s needed. How can we shift investment and political incentives to truly enable the proliferation of sustainable technology worldwide?

Silvia Weko

Can digital be the great global leveler?

Deep-rooted digital divides within and between countries are currently increasing inequalities, with huge disparities in access to technology, infrastructure, and digital literacy. How can the proposed Global Digital Compact create a more inclusive and equitable digital landscape?

Carlos Maria Correa