Members of the Cornell community gathered in the student lounge on October 20 for a wide-ranging discussion of Data Access and AI Explainability, the newest book by Frank Pasquale, professor of law at Cornell Tech and Cornell Law School and a leading scholar of the law of artificial intelligence. Pasquale was joined by panelists Karen Levy, associate professor in the Department of Information Science and associate faculty member at Cornell Law School; Jessica Eaglin, professor of law; and Steve Jackson, professor in the Department of Information Science.

Steve Jackson (left) and Frank Pasquale

Pasquale opened the conversation by situating the book in more than a decade of work on data protection, algorithmic scoring, and technological due process. Reflecting on the origins of the project, he explained that “people are increasingly being judged algorithmically.” Sometimes these algorithms are faulty, including when they are based on inaccurate or inappropriate data. Pasquale explained that a core principle animating his work is that “if you’re being scored, you ought to know what data was used for the scoring, and know how it was processed.” He also mentioned that, in some situations, scored persons should have an opportunity to “be an exception to the rule,” making their case with narratives rather than quantitative data.

Levy emphasized the sociological depth of Pasquale’s account and its contribution to debates about data access and explainability. “I really think he is a real north star for what we ought to aspire to be doing in law and technology scholarship,” she said. Highlighting a passage she described as central, Levy quoted Pasquale’s argument that “the net effect of individual information access rights is to build a new form of intellectual infrastructure for efforts to address unfair power differentials in the data economy.”

Karen Levy (left) and Jessica Eaglin

Eaglin, whose work examines the use of algorithms in criminal law, underscored the book’s insistence that transparency is only a starting point for political and social engagement with automated systems. “Transparency is a starting point for so much more,” she noted, adding that Pasquale’s project “is really bringing that forward” by showing how data accessibility “provides a basis for collective social action.”

Jackson praised the book’s clarity and analytical precision, observing that it “is a short, clear book,” and one he had already shared with students. He also emphasized that Pasquale’s typology of problematic data—“inaccurate, inappropriate, and discriminatory”—provides an important framework for understanding algorithmic harms.

The discussion concluded with an extended Q&A, during which Pasquale reflected on explainability, regulatory design, and the challenges posed by emerging generative systems.

Pasquale’s Data Access and AI Explainability is available open access on Cambridge Core.