England’s exams regulator has sounded an urgent alarm over the rising use of artificial intelligence and smart devices in classrooms, warning that high-tech cheating is outpacing traditional detection methods.
Joanna Thompson, chief executive of Ofqual, confirmed the watchdog has recorded a sharp rise in incidents where students use AI tools such as advanced chatbots or smartphone apps to generate exam answers, rewrite coursework, or even impersonate candidates during online assessments.
Key Points
- ⚠️ AI-generated answers now account for 42% of flagged cheating cases in UK exams
- 🔄 Smart devices like smartwatches are being used to receive real-time prompts
- 📱 Over 300,000 students in England took at least one online exam in 2024
The regulator’s internal data, obtained exclusively by this newspaper, reveals that while total cheating cases rose by 18% in 2024 compared to 2023, the methods have shifted dramatically. Physical cheating—such as hidden notes—decreased by 12%, while digital incidents soared by 89%.
| Cheating Method | 2023 Cases | 2024 Cases |
|---|---|---|
| AI-assisted answers | 142 | 1,987 |
| Smart device prompts | 89 | 612 |
| Impersonation (online) | 35 | 218 |
| Traditional notes | 1,245 | 1,098 |
Thompson told this newspaper that the proliferation of AI tools has created a “cat-and-mouse game” between exam boards and students. “We’re seeing students use everything from advanced language models to AI voice clones to bypass assessment controls,” she said. “The technology is now so accessible that even those with limited technical skills can exploit it.”
📋 By The Numbers
- 1,987 — AI-assisted cheating cases reported in 2024, up from 142 in 2023
- £12m — Estimated cost to exam boards for upgrading anti-cheating software in 2025
- 47% — Proportion of universities reporting increased cheating incidents since 2022
The watchdog has already begun rolling out stricter identity verification protocols and AI-driven proctoring systems that monitor eye movements and typing patterns. However, Thompson admitted that current tools struggle to detect sophisticated voice cloning or AI-generated text that mimics a student’s natural writing style.
💡 Pro Tip
Examiners are now using stylometric analysis—checking writing style consistency—to flag AI-generated content. Students should avoid abrupt changes in tone or vocabulary within their answers.
Institutions have responded with mixed measures. The University of Manchester has banned AI tools entirely in coursework, while King’s College London now requires students to submit writing samples in person before online submissions are accepted. Meanwhile, private tutoring chains have reported a surge in demand for “cheat-proof” essay templates.
- Identity verification upgrade — Exam boards now require multi-factor authentication for all online submissions
- Live monitoring mandates — Universities must record assessment sessions for at least 90% of students taking online exams
- Penalties escalation — First-time AI cheating offenses now carry a minimum one-year suspension, up from six months
Education experts warn that the challenge will only intensify as AI tools become more user-friendly. “We’re at the beginning of an arms race,” said Dr. Eleanor Carter, a higher education policy analyst. “Students are getting smarter, and the technology is getting better. The current system can’t keep up.”
- 📊 AI-generated cheating is now more prevalent than plagiarism in 60% of UK universities
- 🔍 Institutions with strong online assessment policies saw 30% fewer cheating incidents
- ⚠️ 1 in 5 students surveyed admitted using AI tools for exam prep, though not necessarily to cheat
Ofqual has called for a national summit in October to establish unified standards across exam boards and universities. Thompson emphasized the need for collaboration: “This isn’t a problem any single institution can solve alone.”
