Artificial intelligence (AI) and human development may appear to be opposing forces. One rooted in the synthetic, the other in the very essence of humanity. Yet, AI holds the potential to reimagine and actualize what human development can achieve in the coming decades. This opportunity lies not in seeing AI merely as a tool for development or a risk to govern, but as a new sphere for broadening human choices and agency. As presented by the UNDP’s Human Development Report 2025, launched this month, the goal is to make AI work for human development, enhancing our capabilities rather than imitating, replacing or pretending to be human. The true promise of AI lies not in its novelty, but in how meaningfully it can contribute to solving the problems that shape our modern lives.

In the Asia Pacific region, as it is globally, the discourse on AI has been polarized. On one hand, AI has been heralded as the ultimate positive disruptor, capable of offering a plethora of innovations and accelerating change that can shake and stir progress at scale. On the other hand, it has been portrayed as an existential threat, with fears of massive job losses, societal upheaval, increased income and gender gaps, and heightened energy use and hence environmental pressures. While both perspectives hold some validity, they may miss the moment afforded to us  to guide and govern AI as a fundamental enabler of human development.

At UNDP, working closely with partners, we are embedding AI in human development approaches and investments to make it generative, or "GenHD". Here are some early steps as we embark on this journey of discovery.

• AI that expands human choices. Testing AI applications that can revolutionize the delivery of essential public services, such as education, healthcare, and social protection which can provide greater access, with less physical infrastructure requirements. By making these services more efficient and accessible, it helps reduce inequalities. By initiating AI for social protection in Malaysia, from policy adjustments to data systems and targeting, the intention is to support government make social protection delivery better targeted, faster to reach end user, and be rapidly adjusted to changing situations especially during emergencies.

• AI to minimize risk: Prevention is one of AI’s most powerful applications. From early-warning systems to risk assessments, AI can help individuals, communities, and governments make more informed decisions. In healthcare, AI applications are rapidly reading diagnostic tests and enabling early interventions. Predictive AI tools are helping farmers adapt earlier to changing weather patterns, improving crop yields and food security.  In India, the demonstrated use of AI enabled geospatial datasets helps with smart agriculture, from estimating crop yields to anticipatory  pest control, guiding farmers in their planting, irrigation, and harvesting techniques.

• AI that goes local: One of the greatest challenges of AI is its reliance on global models that often overlook local languages, intelligence, and cultural practices. AI is figuring out that local knowledge is essential for addressing uniquely contextual challenges and must adapt to the specific needs of local communities. In Indonesia, for example, an AI application for participatory village planning has shown its value for local governance solutions across 100 different villages and attempts to address each unique  circumstance.

• AI to build trust: Trust is fundamental to good governance and social cohesion. AI applications that strengthen accountability and transparency has shown to deepen trust between citizens and governments. This includes generating data and dialogue as part of continuing public discourse, while reflecting and respecting diverse viewpoints, and promoting inclusivity. A new chatbot in Nepal allows citizens to better understand the public services available and takes in peoples’ feedback to improve service delivery. In Sri Lanka, the government is integrating AI into public services while refining AI governance and ethics to improve citizen trust.

When the concept of human development was introduced, it incorporated three key dimensions: health, education, and life expectancy as measures of human progress. All three aspects can be impacted by AI. If used with this intent. By aligning AI with the principles of human development, we shape AI applications that are less synthetic. And yes, more  humane.