South Korea will send a nationwide phone alert on Thursday morning as more than 500,000 students sit the college entrance exam. The roughly 10-second alert will remind people to help detect hidden devices. Proctors will inspect glasses and seal shirt buttons with stickers, while an additional 8,400 supervisors will monitor the exam, AP reported.
The measures follow a case in which a top scorer in a national tax exam revealed he had used AI to get help. Similar concerns are prompting changes to assessments in other countries.
In England, the Office for Students said last week that AI had made it “close to impossible” to assess university coursework. It urged institutions to return to traditional exams and in-person presentations, The Guardian reported. Josh Fleming, an Office for Students director, called AI “a generational risk to the system as it stands.” A Guardian investigation found almost 7,000 proven cases of AI cheating at UK universities in 2023-24, up from 1.6 cases per 1,000 students the previous year.
During China’s gaokao in June 2025, when more than 13 million students took the national entrance exam, major AI apps suspended photo and question-answering tools during test hours. ByteDance’s Doubao told a user who uploaded an exam question that “the question-answering service will be suspended,” according to The Guardian.
Some professors in the United States are adding oral defenses to written work. Chris Schaffer, a biomedical engineering professor at Cornell, said: “You won’t be able to AI your way through an oral exam!” He now holds 20-minute oral defenses after written problem sets. At the University of Pennsylvania, a teaching center director also described a shift toward in-person assessments.
AI detectors have struggled to keep pace with tools that rewrite papers and imitate human typing. As institutions adopt measures including in-person and oral exams, stickers on clothing and phone alerts, software has not stopped cheating, the article says.
