The grade you get might not accurately describe what you've learned
In England's examination halls, a quiet arms race is unfolding between the institutions that certify human knowledge and the technologies that allow that knowledge to be borrowed, fabricated, or smuggled in. Ofqual, the body charged with safeguarding the integrity of national qualifications, has begun to reckon openly with the possibility that the tools of cheating are now evolving faster than the rules designed to prevent it. At stake is not merely fairness between individual students, but the shared social trust that makes a grade mean something — the belief that a credential reflects what a person actually knows.
- Record numbers of students were caught using mobile phones and smartwatches during last summer's GCSE and A-level exams, signalling that device-based cheating has become the dominant threat to exam security.
- Smartglasses that project invisible text and near-undetectable earpieces are already being marketed, threatening to make the next wave of cheating essentially impossible to spot in an exam hall.
- AI-generated coursework is quietly eroding another pillar of assessment, with teachers reporting that distinguishing human writing from machine output grows harder with every passing term.
- Regulators are weighing drastic responses — including abolishing coursework altogether — as softer interventions like increased check-ins and citation requirements struggle to keep pace with the scale of the problem.
- The credibility of England's entire qualification system hangs in the balance: if grades can no longer be trusted, universities, employers, and the students who earned their marks honestly all lose something that cannot easily be restored.
Ian Bauckham, who leads Ofqual, England's qualifications watchdog, recently gave voice to a concern that has been building quietly inside the exam system: the technology students use to cheat is outpacing the rules designed to catch them.
Last summer, Ofqual recorded 2,225 cases of students found with mobile phones or smartwatches during GCSE and A-level exams — the largest single category of cheating the regulator has tracked every year since 2018. But Bauckham's deeper worry is what comes next. Smartglasses that project text only the wearer can see. Earpieces small enough to be nearly invisible. Fully internet-connected smartwatches. All of it already available, already being advertised to students.
The consequences are not merely procedural. When a grade no longer reflects what a student actually knows, it becomes unreliable — and if enough grades are unreliable, the entire qualification system loses its meaning. Bauckham described England's exam infrastructure as "a real national asset," one whose value depends entirely on the trust placed in it by universities, employers, and the public.
Coursework presents a separate but equally urgent problem. Teachers across the country have told Bauckham that AI-generated writing is becoming harder and harder to identify, leaving the authenticity of submitted work in genuine doubt. Ofqual is now weighing its options as it redesigns GCSEs and A-levels: eliminate coursework entirely — what Bauckham called the "nuclear option" — or introduce more rigorous safeguards, such as requiring students to explain their thinking at multiple stages, or to cite sources in enough detail that simply pasting in AI output becomes impractical.
The regulator's central difficulty is one of timing. Technology evolves faster than policy. By the time new rules are written, tested, and implemented, the devices they target may already be obsolete. Bauckham acknowledged the pressure plainly, saying Ofqual needs to act "really fast" — though whether fast enough to stay ahead of those determined to game the system remains, for now, an open question.
Ian Bauckham runs the Office of Qualifications and Examinations Regulation, the body that oversees England's school exams. Recently, he sat down for a podcast and said something that should worry anyone who cares whether grades mean anything: the technology students are using to cheat is getting faster and smaller and harder to catch than the rules meant to stop them.
Last summer, Ofqual counted 2,225 cases of students caught with mobile phones or smartwatches during GCSE, AS, and A-level exams. That's the biggest category of cheating the regulator has tracked every single year since 2018. But Bauckham wasn't talking about phones. He was talking about what comes next. Smartglasses that project text only the wearer can see. Earpieces so small they're nearly invisible. Smartwatches that are fully internet-connected. All of it already being advertised. All of it already in the pipeline.
"Technology is changing fast," Bauckham said on the podcast. "We are all familiar with mobile phones but there are smartwatches that we are increasingly seeing on young people that are fully internet connected and so present many of the same challenges as mobile phones." He went on to describe smartglasses that would play text across the inside of the lens—visible only to the student wearing them. "We are going to have to keep on top of this," he said.
The stakes are not abstract. When a student gets marks they didn't earn, the grade they receive stops being a reliable measure of what they actually know. It becomes a lie. And if enough students cheat, the entire qualification system—which Bauckham called "a real national asset"—loses its value. Universities won't trust the grades. Employers won't trust them. The whole point of having standardized exams collapses.
But the phone problem is only half the story. Teachers are also struggling to detect when students submit work generated by AI. Bauckham said he's been told repeatedly by educators that AI-generated content is "getting harder and harder" to spot. So Ofqual is now asking a hard question as it redesigns GCSEs and A-levels: should coursework even exist anymore? Or if it does, what safeguards would actually work?
One option is the "nuclear option," as Bauckham put it—eliminate coursework entirely. Another is to require teachers to check in with students more frequently as they work, asking them to explain their thinking and sign off on authenticity at multiple points. A third is to demand that students cite their sources and explain their reading in detail, making it much harder to simply paste in whatever ChatGPT produces.
"What you haven't done is just open ChatGPT and say: 'Write me 10,000 words on Henry VIII's foreign policy, please,' because that is absolutely not acceptable," Bauckham said. The comment was pointed. It was also a reminder that this is exactly what some students are doing right now, and the system has no reliable way to stop them.
Ofqual faces a timing problem. Technology moves faster than regulation. By the time new rules are written and implemented, the devices they're meant to catch will already be obsolete, replaced by something smaller, smarter, harder to detect. Bauckham knows this. He said the regulator has to act "really fast." But fast enough to stay ahead of the people trying to game the system? That's the question that keeps him up at night.
Citações Notáveis
Our qualification system is a real national asset and we have to keep on top of this to stop this national asset being undermined— Ian Bauckham, head of Ofqual
If you gain help unfairly on a mobile phone or a smartwatch or any other kind of device, you are potentially getting marks in the exam that you don't deserve— Ian Bauckham
A Conversa do Hearth Outra perspectiva sobre a história
Why does it matter if a student uses a smartwatch to look up an answer during an exam? They're still the one taking the test.
Because the grade is supposed to measure what they know, not what they can find in thirty seconds. If you can't trust the grade, you can't trust the qualification. Universities won't know if an A really means mastery or just access to Google.
But teachers have always caught cheaters. Why is this different?
Scale and invisibility. A phone is obvious. A smartglass that projects text only you can see? A teacher standing at the front of the room has no way to know it's there. And there are thousands of students taking exams on the same day.
What about the AI coursework problem? Can't teachers just ask students to explain their work?
They can try. But if a student has already submitted ten thousand words generated by ChatGPT, asking them to explain it after the fact is almost useless. They can bluff their way through. The real solution would have to catch it before submission.
So Ofqual might just ban coursework altogether?
That's one option they're considering. It's drastic, but it solves the problem cleanly. No coursework, no AI to detect. But it also means losing a form of assessment that can measure things exams can't—like sustained research and writing over time.
How much time does Ofqual have to figure this out?
Not much. The technology is already here. Students are already using it. The regulator is playing catch-up, and catch-up is a losing game when the other side has a new innovation every six months.