The role of artificial intelligence (AI) in higher education often appears troubling, with many students using AI tools to cheat on assessments and open-book online exams, seemingly diminishing genuine critical thinking. The upcoming graduates might finish degrees without truly engaging in deep analysis. Personally, I avoid using ChatGPT due to my course's closed-book exams and environmental concerns over AI data centers, though students generally accept AI as a learning aid. While debates emphasize “cheating, ” AI increasingly supports research and essay structuring. Concerns about misuse of large language models (LLMs) in education are valid, but understanding their rising use requires examining the context that led here. In March 2020, at age 15, I and many students welcomed school closures caused by COVID-19 lockdowns, initially thinking it was a brief break. Instead, my education faced major disruption over three years. GCSEs and A-levels were canceled and replaced with teacher-assessed grades favoring already high-performing private schools. Further closures and indecision led to another cancellation in 2021. My 2023 A-level cohort was the first to encounter “normal” exams again, triggering strict anti-grade-inflation measures leaving many students with disappointing results. Universities, too, struggled to assess students off-campus, resorting to open-book, online exams where no coursework existed. Even five years later, 70% of UK universities still employ some form of online assessment.

This shift doesn’t reflect a decline in standards; rather, most recent students missed full national exam experiences and key curriculum due to prolonged closures and changing exam formats. This continual government indecision fostered uncertainty still influencing higher education assessments. In my university experience, half of first-year exams were online, while this year, exams returned fully to handwritten, closed-book formats—often with confirmation of exam format delayed well into the academic year. Third-year peers took identical exams online with longer time allowances, acknowledging their lack of handwritten exam experience during their degrees. When ChatGPT was launched in 2022, it arrived amid this unsettled university environment, marked by inconsistent and varied exam formats across institutions and faculties. Such inconsistency heightened temptation for students feeling disadvantaged and complicated AI usage detection. Apart from flawed exams, the student experience is more financially challenging than ever: 68% hold part-time jobs, a decade-high figure, while student loan burdens fall heaviest on the poorest. I belong to the first cohort repaying loans over 40 years instead of 30—with tuition fees expected to rise again. Students thus have less time to fully dedicate to studies. AI functions as a time-saving tool; if students can’t engage deeply, flaws lie within the university system itself. AI use is booming not just because it is fast and convenient, but also due to lingering post-COVID exam uncertainties and increasing student financial precarity. Universities must settle on consistent exam formats. If involving coursework or open-book tests, clear guidelines on “proportionate” AI use are essential. AI is here to stay—not because students are lazy, but because the student experience is evolving as rapidly as technology itself.