Skip to content
Unite.AI·Indexed by this site · Originally published· Governance →· Its place in the scan record →

England's exam regulator Ofqual plans to adjust A-Level assessment methods due to AI misuse risks

Ofqual、AIの不正利用リスクを受けてAレベルの評価方法の変更を計画

AI summary

England's exam regulator Ofqual proposed adjusting the non-exam assessment methods for A-Level History and three English subjects to address authenticity risks brought by AI. In a letter published on 18 November 2026, regulator head Sir Ian Bauckham said it is increasingly difficult to confirm whether long-form writing completed without supervision is entirely the student's own work, and that if the authenticity of the work cannot be assured, changes to the assessment itself need to be considered. The letter also requires exam venue staff to receive training before the summer 2027 exam season on identifying and responding to misuse of wearable devices, and awarding organizations must reply by 30 November 2026 explaining risk assessments and preventive measures. A previous letter's statistics showed about 5,000 reported cheating cases each year over the past three years, nearly half involving bringing communication devices such as mobile phones into exam venues; penalties in 2025 included 1,125 cases of students losing all GCSE or A-Level qualifications and about 2,000 cases of mark deductions; confirmed AI cheating cases at the time were relatively few, but teachers and school leaders were very concerned about the actual scope of misuse.

Read originalunite.ai