Higher education institutions redesign assessments amid rise of generative AI

Generative artificial intelligence is forcing a fundamental redesign of academic assessments and teaching methodologies in higher education. As AI becomes capable of completing most undergraduate assignments, institutions are moving away from evaluating only finished products, such as essays or code, toward more holistic evidence of learning.

At the Massachusetts Institute of Technology (MIT), a dedicated committee has called for a broad review of curricula, suggesting that instructors reconsider learning goals. Recommendations include increasing the use of oral exams, portfolios, and in-person discussions to verify student understanding. MIT President Sally Kornbluth described the technology's impact as a “watershed” moment for the university.

Educators are increasingly focusing on academic integrity through “ownership tests,” where students must demonstrate they can explain, trace, and challenge the intellectual decisions made in their work. Rather than implementing blanket bans, which are difficult to enforce, many are adopting nuanced policies that permit AI for brainstorming or research while prohibiting its use for final analysis or composition. This shift aims to ensure that students continue to develop critical thinking and reasoning skills that might otherwise be bypassed by automated tools.