AI Singapore and US National Science Foundation Launch Joint Research on Trustworthy AI

AI Singapore (AISG) has launched a joint research project with the US National Science Foundation, bringing together researchers from Nanyang Technological University, Singapore Management University, and the Massachusetts Institute of Technology. The initiative targets a fundamental challenge in AI: how to make human-AI collaboration systems robust and trustworthy when they're tasked with complex, combinatorial decision-making problems.

Current AI systems like ChatGPT excel at conversational tasks but struggle with structured optimisation problems where multiple constraints and trade-offs need to be navigated. The project aims to establish natural language as a reliable interface for what's called "trustworthy AI-human interactive optimisation" — essentially, letting humans guide AI through complex decisions using plain language while keeping the system verifiable and aligned with operational requirements.

The Singapore team is led by Professor Jie Zhang of NTU as principal investigator, with Assistant Professor Zhiguang Cao of SMU as co-PI. MIT's team includes Associate Professors Jacob Andreas and Cathy Wu. The research spans both theoretical foundations and practical deployment considerations, contributing to what both nations view as strategic AI competitiveness priorities.

Why it matters for Singapore: This collaboration puts Singapore-based researchers at the cutting edge of AI alignment and human-AI interaction — a field that will define how AI gets deployed in high-stakes settings like government services, healthcare, and financial regulation. For Singapore's Smart Nation ambitions, trustworthy AI isn't just an academic concern; it's a prerequisite for public-sector AI adoption. The NTU-SMU-MIT pipeline also strengthens Singapore's position as a research bridge between Asia and top US institutions, which matters for talent attraction and research funding flows.