A research team from Shahjalal University of Science and Technology has won a featured presentation at the International Conference on Machine Learning (ICML), one of the world’s most competitive forums in artificial intelligence.
The paper, “Position: AI governance requires not just laws but also interoperability protocols like ISO,” will be presented on May 11 in Seoul, South Korea.
The authors of this paper are Azmin Tusik Wasi, Rafia Islam, Mahfuz Ahmed Anik, Mond Manjurul Ahsan, Taki Hasan Rafi, and Donkyu Choi.
The study highlights the fragmentation of global AI regulation, noting that frameworks such as EU AI law and the NIST AI Risk Management Framework apply different standards for safety and compliance.
To address this, the authors propose an ISO-like interoperability protocol and introduce the concept of “AI nutrition labels.” This is a machine-readable summary that captures metrics such as bias, energy usage, and data provenance. This approach aims to streamline the cross-border evaluation of AI systems and reduce duplication of regulatory processes.
The team represents multiple institutions, including Bangladesh Independent University Soust and Hanyang University, with Dr. Manjulul Ahsan serving as research supervisor.
Spotlight papers at ICML are selected for their potential impact, and this is one of the few from Bangladesh to receive such recognition, making it the first position paper to receive this honor from Sust.
