Leading university presidents, including at Dartmouth and Harvard, argue that AI can accelerate knowledge creation and should be used productively and ethically to keep education and research at the forefront.

Across universities, leaders generally embrace AI as essential for preparing students to work with AI and to stay relevant in a rapidly evolving landscape.

Meanwhile, some K-12 systems are pulling back on generative AI, with New York City suspending its use for grades K-8 for a year and Los Angeles Unified also tightening restrictions.

Policy fragmentation persists, as many institutions leave AI-use standards to individual professors, creating confusion about what counts as cheating.

Institutions are enacting various rules, from the University of Chicago limiting AI in certain social science courses and Chicago Law restricting device use in first-year classes, to UC Berkeley Law banning AI for brainstorming or drafting graded work, and Cornell updating disciplinary procedures for AI-related misconduct.

Other actions include Ohio State mandating AI training across majors to ensure fluency, while Harvard is exploring guidelines to deepen learning with AI without compromising integrity.

A large study in China found AI-enhanced homework improved scores, but control-group tests suffered, suggesting deep problem-solving may decline when AI is overused for homework.

The broader context includes ubiquitous AI tools on platforms like Google Classroom, underscoring the need for clear guidelines on acceptable use and misuse detection.

Administrators advocate thoughtful, ethical integration and responsible AI education, with some calling for in-class testing and AI-proof assignments to curb misuse.

The New York Times highlights policy variability and the ongoing challenge of defining misconduct, fueling calls for responsible AI governance.

Overall, higher ed is integrating AI education with boundaries to preserve learning integrity and critical thinking.

MIT research indicates AI use can shift campus culture toward 'cognitive surrender,' with students relying on bots for quick answers and potentially hindering memory and deep thinking.