Nikola Banovic receives NSF CAREER Award to advance explainable AI

Prof. Banovic aims to use human-AI interaction to explain and justify AI decisions to end users.
Prof. Nikola Banovic
Prof. Nikola Banovic

Nikola Banovic, assistant professor of computer science and engineering at the University of Michigan, has received a National Science Foundation (NSF) CAREER Award in support of his research to democratize artificial intelligence (AI) by delivering explanations of AI decisions to end users. His project is titled “Achieving Explainable Artificial Intelligence (AI) through Human-AI Interaction.”

The NSF CAREER Award is a highly competitive honor, given to only 500 early-career faculty across all science and engineering disciplines each year. The award supports young scientists who “have the potential to serve as academic role models in research and education and to lead advances in the mission of their department of organization.”

The recent surge in AI-based technology has sparked broad concerns about AI trustworthiness as well as its accessibility to non-experts. To address these issues, Prof. Banovic seeks to use carefully designed human-AI interactions grounded in sensemaking theory to provide explanations of AI models’ decision-making to end-users without sacrificing AI performance. His goal is to provide users with information about how AI makes decisions as well as its overall decision-making capabilities.

This advancement of explainable AI is critical to building societal trust in and adoption of AI-based technologies, as well as ensuring that we understand AI’s limits in high-stakes decision-making scenarios, such as in healthcare, government, law, and beyond.

Prof. Banovic’s research at U-M has already made significant headway in enhancing the explainability and interpretability of AI through human-computer interaction, as well as promoting ethical and responsible AI applications. His work has explored how laypeople understand and perceive uncertainty in AI decision-making, exposed errors in online government screening tools, revealed the ways in which untrustworthy AI can deceive end users, and more.

His CAREER-supported research will contribute further to the democratization of AI, making AI’s decision-making processes and abilities more transparent and understandable to everyday users.

Explore:
Honors and Awards; Nikola Banovic; e-HAIL