Over the past decade, law school exams have steadily expanded beyond the traditional in-class, closed-book model. Driven by ABA guidance, increased accommodations, and evolving pedagogy, schools adopted open-book, take-home, and hybrid formats at scale. The COVID period accelerated that shift, normalizing flexible, resource-rich exam environments supported by reliable take-home systems.
That model depended on a stable assumption: while students could access materials, they could not easily access assistance. The emergence of generative AI in late 2022 disrupted that assumption. Suddenly, any environment that allowed communication, whether with online services, local tools, or other people, became inherently insecure. The response across law schools was immediate: a rapid return to in-class, closed exams.
However, this reversion exposed a structural conflict. Schools still require students to engage with statutes, casebooks, and notes as part of modern legal education. Simply removing access to resources is not pedagogically viable. The challenge, therefore, is not choosing between open and closed exams, but enabling necessary materials while preventing communication.
This session presents a practical framework for resolving that conflict through controlled access. We identify four categories of information that students may legitimately need during an exam: institution-controlled exam content, student-selected local materials such as notes, the open internet, and externally hosted but controlled resources. We examine why the open internet, even when filtered or whitelisted, cannot reliably meet exam security requirements, and we outline an alternative model based on tightly constrained environments.
In particular, we introduce an exam-safe content platform that allows students to access casebooks and annotations during a closed exam without exposing communication pathways. The goal is not to eliminate flexibility, but to redefine it within enforceable boundaries. We will also invite discussion on whether this framework fully captures the space of permissible materials and where institutions see remaining gaps or edge cases.