RTX BBN Technologies Develops AI for Regulatory Compliance

RTX BBN Technologies has been awarded a contract by the Defense Advanced Research Projects Agency (DARPA) to support its Human-AI Communication for Deontic Reasoning Devops (CODORD) program, announced in a press release. The initiative aims to bridge the gap between human regulations and AI comprehension, enabling computers to interpret and apply regulations in a verifiable and explainable manner.

The project, named DONATELLO (DeONtic AssistanT Enabled by Large Language mOdels), is a human-in-the-loop AI solution designed to simplify and translate complex regulatory texts into executable logic. This approach addresses the challenges posed by regulatory texts, which are often filled with exceptions and nuances that traditional programming struggles to handle.

Dr. Brian Ulicny, BBN's principal investigator, emphasized the need for systems that can explain why specific actions are permitted or prohibited in ambiguous situations. The BBN team, which includes Johns Hopkins University and Stony Brook University, will conduct their work in Cambridge, Massachusetts; Baltimore, Maryland; and Stony Brook, New York.