Designing Responsible AI Technologies to Curb Disinformation

Our project investigates two key societal challenges. First, how can we design responsible AI technologies to curb the digital spread of disinformation? Second, by grounding research on responsible AI in the context of a real societal challenge – disinformation – what broader lessons can we learn about how to better design, build, and evaluate responsible AI technologies in general? For example, while there is vast technological research on creating explainable AI models, much of this research today lacks a real, practical use case and meaningful evaluation. Moreover, because disinformation often targets and disproportionately impacts different demographic groups, how can we measure and ensure the fairness of AI models for prioritizing human effort in fact checking? The problem of disinformation thus offers an invaluable, concrete testbed for grounding broader research in developing and testing fair and explainable AI models.