MIT and Singapore Agency Fund 13 Projects to Advance Trustworthy AI
MIT Schwarzman College of Computing and Singapore's Defense Science and Technology Agency provide funding for 13 new artificial intelligence research projects. The initiatives focus on building trustworthy AI, enhancing human cognition, and making AI accessible to everyone.
The MIT Schwarzman College of Computing and the Defense Science and Technology Agency (DSTA) in Singapore award funding to 13 new artificial intelligence research projects. This collaboration aims to advance AI technologies that build trust, counter asymmetric threats, and adapt to complex environments. Each selected project targets at least one of three core themes: trustworthy AI, enhancing human cognition in complex environments, and AI for everyone.
Among the funded initiatives is a project led by Professor Aleksander Madry called SYNTHBOX, which focuses on establishing real-world model robustness and explainability. Madry’s team leverages recent advances in realistic graphics rendering and data modeling to build a new toolbox for the streamlined development of trustworthy machine learning solutions. Another notable project comes from professors Regina Barzilay and Tommi Jaakkola, who are creating next-generation natural language processing technologies for low-resource tasks.
Barzilay and Jaakkola address the challenge that most languages in the world lack rich annotation, which often leads to inaccurate and brittle AI outputs. Their team develops new text-generation tools for controlled style transfer and novel algorithms to detect misinformation in these under-resourced languages. Together, all 13 projects represent a significant step forward in making AI systems more reliable, transparent, and globally inclusive.