AI is breaking how schools test students and what they choose to study
Global exam scandals tied to AI cheating are colliding with a wave of US students switching majors out of fear AI will wipe out their future jobs.
What happened: This summer, AI collided with education on two fronts. Exam season turned into unrest worldwide: Mexico forced nearly 60,000 university applicants to retake entrance tests over suspected mass cheating, Portugal's rushed digital marking system produced errors that sparked protests demanding the education minister's resignation, and in India a leaked exam paper affected millions of students, was linked to more than a dozen suicides, and forced the education minister to quit. Meanwhile in the US, a new survey found 69% of college students worry AI will make it harder to land a job after graduation, and 22% say they've already switched majors because of it.
Why it matters: AI has made two pillars of education shaky at once: the exam, as proof of learning, and the college major, as a bet on future employability. When cheating scandals and marking failures collide with an already-precarious job market, students lose faith that grades reflect real effort or fairness, says academic-integrity researcher Sarah Eaton. That broken trust has spilled into politics, costing two education ministers their jobs and feeding a wider youth protest movement in India over jobs and opportunity. On campus, uncertainty about which skills will still matter is pushing one in five US students to abandon majors they'd already chosen.
Two problems, one root cause: On the testing side, generative AI can write a passable essay in seconds, so essays no longer reliably show what a student actually knows, while tech meant to stop cheating, like Portugal's digitised marking, has itself failed or been gamed. On the career side, AI is absorbing entry-level tasks in fields like accounting, so employers now expect graduates to also manage AI tools, present findings and defend their own reasoning aloud, skills traditional majors weren't built around. Neither problem has a real fix yet: assessment needs rebuilding, and colleges are only starting to bolt AI-literacy and critical-thinking training onto older curricula.
The rollout: Denmark now requires students to verbally defend written essays starting this term, and Calgary professor Sarah Eaton has adopted similar spoken defenses in her own classes, arguing the traditional essay should "die with dignity." American University's business school pairs AI-skills training with communication and teamwork requirements, and OpenAI's education VP says critical thinking and AI literacy, not any single major, will be what employers reward. But these remain patchwork fixes: most universities and governments are reacting case by case, with no shared standard for testing fairly or preparing students for a job market still shifting under them.