Clear, Compelling Arguments: Rethinking the Foundations of Frontier AI Safety Cases

Shaun Feakins, Ibrahim Habli, Phillip Morgan

This paper contributes to the nascent debate around safety cases for frontier AI systems. Safety cases are structured, defensible arguments that a system is acceptably safe to depl…