Three ways that AI is supporting policymakers in using geospatial data

When artificial intelligence (AI) meets geospatial data, it was more than just optimising workflows. 

Speakers at the “AI to Enable the Potential of Geospatial Data for Government Products” panel at Geo Autonomy Summit 2026 in Singapore on March 31 highlighted how this convergence is creating a more responsive, inclusive and data-driven foundation for the future of public sector.  

Instead of being just confined to static maps, today’s geospatial data is evolving into a dynamic system of “living models” powered by AI automation, digital twins, and predictive modelling. 

A “single view” of massive and diverse datasets also makes collaboration among different stakeholders easier, including those who might not be geospatial experts. 

Singapore Land Authority (SLA)’s Director for Survey and Geomatics, Victor Khoo and Asian Development Bank (ADB)’s Principal IT Specialist (Technology Innovation) Marc Lepage highlighted how this tech synergy is transforming data from a backend resource to a bridge for trust and localised action. 

Turn massive, raw data into usable maps and models 

SLA’s Khoo noted that machine learning (ML) and deep learning have been key to turning raw geospatial data into clean, usable layers for various agencies to use in their decision-making process.  

These layers included a range of models and information.  

Today, national mapping now depends on many heterogeneous data sources, including the private sector’s, he shared. 

This was why it was so important for agencies like his to maintain formats and standards to effectively combine the sources and produce reliable, consistent national data sets.  

Highlighting that there was “no room for error,” Khoo emphasised the importance of agencies getting formats, specifications, and data quality right at the upstream stage to prevent AI from processing bad inputs and propagating the errors. 

He also added that while SLA is integrating more AI-driven automation and prediction into its workflows, it remains cautious about relying too heavily on prediction when the standard for official mapping requires verifiable accuracy. 

The panel in which Khoo spoke also championed a public-private partnership in leveraging AI and the best of private sector’s tech on top of robust data foundations. 

Leverage synergy of digital twins to simulate and predict governance 

ADB’s Lepage said that AI could be used on top of geospatial layers to model risk, forecast scenarios, and support digital twins.  

The use cases for governments ranged from managing disaster risk, environmental impact and smart city operations. 

He highlighted two of ADB's working products, which included a disaster risk platform that uses a combination of AI, ML, generative AI (GenAI) on geospatial data to simulate and communicate the risks. 

The platform, first made available to government decision-makers in the Pacific region, was progressively rolled out to other organisations working with the local communities. 

Lepage highlighted the potential of also using the platform to engage citizens to promote understanding of the risks and implications.  

The other product was a digital twin of a 25km river in Southeast Asia to simulate development scenarios amidst plastic pollution.  

The project covered 55 government agencies, with the ADB developing a customised digital twin proof-of-concept (POC) for each to meet their unique needs.  

The combination of AI and digital twin has provided ADB with "a single view of a very complex system” to engage with diverse stakeholders, be it citizens or decision-makers. 

This capability has also helped ADB address both the localised needs of the agencies and the larger-scale crisis of plastic pollution. 

Make geospatial intelligence understandable with natural language 

GenAI serves as an accessibility layer on top of geospatial data for non-technical experts and citizens, through its ability to allow users to ask questions and explore scenarios around the data. 

Lepage highlights the next trend of using AI on geospatial data to move past top-down governance. This shift would enable more personalised, community-level services that are directly responsive to local needs. 

As data outputs now become explainable with natural language, AI could serve as a bridge for trust and collaboration by supporting broader policy discussions and involving civil society inputs. 

“The technology layer allows us to really fast-track the prototype that quickly gets validation from the citizens or users of the solution. This really allows for better co-design among different partners,” said Lepage.