With less than four years left until the 2030 deadline, the Sustainable Development Goals (SDGs) are far off track. In response, artificial intelligence (AI) is increasingly being cited as a means to accelerate progress. AI can predict flood risk from incomplete hydrological data, model the spread of infectious diseases, optimize agricultural production, predict commodity prices to support the energy transition, and extend legal and financial services to underserved communities. These features are real and well documented.
But at the heart of much of the discussion around “AI for the SDGs” is a fundamental misunderstanding. The challenge is not whether AI can provide technical solutions. What matters is whether governance systems can direct those solutions toward socially desirable outcomes. AI is a tool. It’s not a policy.
This distinction is important. The success of AI applications depends on a set of governance choices regarding rights, responsibilities, accountability, and distribution. Technology can generate predictions, recommendations, and decisions, but it cannot decide who benefits, who bears the risk, or who is responsible if things go wrong.
Capability and governance are not equal
Consider one of the most promising areas for AI implementation: water management. Machine learning systems are currently outperforming many traditional forecasting methods in predicting water demand and managing scarce resources. These systems help utilities improve efficiency, reduce waste, and increase resilience to drought.
Efficiency and equity are not the same, and too often policy frameworks focus solely on the former.
However, the same predictive infrastructure can also be used to allocate access in ways that reinforce existing inequalities. Algorithms can identify who should receive less water in times of water scarcity, but they cannot determine whether that outcome is socially acceptable. Such decisions require rules established through governance.
Issues such as who sets the allocation thresholds, who audits the results, and who has the right to challenge decisions are completely outside the model. Without answers to these questions, increased predictive accuracy and increased social inequity may coexist. Efficiency and equity are not the same, and too often policy frameworks focus solely on the former.
Insufficient explanatory power
Governance challenges become even more apparent when AI intersects with legal systems and public institutions. Predictive policing systems, risk scoring tools, and algorithmic assessments are already used in bail, sentencing, parole, and law enforcement decisions in a variety of jurisdictions. These applications directly impact SDG 16, which aims to promote justice, accountability and strong institutions.
Considerable attention has been paid to making these systems explainable. Modern AI technology now allows users to identify which variables influenced a particular decision or score. Explainability is definitely valuable, but it does not equate to fairness, accuracy, and legitimacy.
Transparency and accountability are related concepts, but they are not interchangeable. Confusing the two is one of the most common governance failures in modern AI policy.
Algorithms can clearly explain why they reached their conclusions, but their inferences are based on biased or incomplete historical data and therefore still produce discriminatory or unwarranted results. The mere fact that a machine can demonstrate a calculation does not satisfy the right to due process. This standard will only be met if individuals have a meaningful opportunity to challenge the decision and an independent body verifies that the system operates within acceptable legal and ethical boundaries.
Transparency and accountability are related concepts, but they are not interchangeable. Confusing the two is one of the most common governance failures in modern AI policy.
Governance gaps that threaten SDG progress
First is the gap between technological innovation and legal liability. Technology evolves much faster than law. Technology standards bodies can issue guidance on self-driving cars, AI-assisted health systems, automated public services, and other emerging technologies long before Congress enacts comprehensive regulations. These standards play an important role in promoting safety and interoperability.
However, standards do not establish legal liability. When autonomous systems cause harm, society must grapple with difficult questions. Does the responsibility lie with the developer who designed the algorithm, the company that introduced it, the hardware manufacturer, or the operator who oversees the system? Only the law can give an authoritative answer.
Policy makers are not forced to choose between enabling and regulating innovation. Rather, we need to recognize that technology and governance issues are inseparable.
Second is the gap between global developments and local impacts. Although AI development is concentrated in a relatively small number of countries and companies, its social and economic impact will be felt locally. Excluding low resource languages indicates a problem. Of the approximately 7,000 languages spoken around the world, only a fraction are meaningfully represented in modern AI systems.
Third, the gap between digital progress and environmental sustainability. The environmental impact of AI is one of the least discussed aspects of the technology’s rapid expansion. Advances in AI rely on energy-hungry data centers, increasingly powerful computing hardware, and global supply chains for semiconductors and electronic components.
Governance of AI for sustainable development
Technological capabilities are advancing faster than the legal, institutional and governance frameworks needed to guide them. As a result, societies are increasingly able to deploy powerful AI systems before establishing clear rules governing their use.
Policy makers are not forced to choose between enabling and regulating innovation. Rather, we need to recognize that technology and governance issues are inseparable. All AI systems deployed to support sustainable development raise questions about accountability, equity, representation, environmental sustainability, and human rights.
If AI is to accelerate progress towards the SDGs, governments must go beyond celebrating technological advances to tackling the harder task of building institutions that can guide AI. The world doesn’t just need AI. We need governance systems to ensure that AI serves the public interest.
This article is based on a keynote address given at City University of New York in June 2026.
