Six in Ten Secondary Teachers Have Spotted AI Misuse

AI in schools has quickly transitioned from a distant strategic consideration into an immediate reality – a tool already shaping how pupils complete work, how teachers assess understanding, and how leaders think about digital governance. 

However, the challenge of digital integrity extends beyond software; it also encompasses a new generation of high-tech hardware, from smart glasses to invisible earpieces, that threatens to undermine the formal assessment process just as AI challenges independent study.

Indeed, recent Teacher Tapp research shows a steady rise in the number of teachers reporting suspected unauthorised AI use by pupils. 

Among secondary teachers, the figure has risen from 15% in April 2023 to 35% in March 2024, 48% in April 2025 and 58% in March 2026. In primary schools, the pattern is smaller but still moving upwards, rising from 1% in 2023 to 8% in 2026.

Secondary schools are already dealing with AI misuse at scale, particularly around independent study, homework and assessment. And, while primary schools may currently feel less exposed, the direction of travel appears set. 

The biggest challenge, perhaps, is that AI is becoming part of the learning environment long before many schools have had the chance to decide what responsible use looks like.

Rather than questioning whether pupils have adopted AI, education leaders must now ensure that staff, pupils, and families recognise when its use is appropriate, when it is harmful, and how it should be managed.

The issue of AI governance

Schools need clarity before they can build confidence

As with the emergence of any new technology, AI has created familiar tensions within school environments: early adopters embrace the opportunities, while others raise valid concerns regarding academic integrity, safeguarding, workload, data privacy and pedagogical impact. 

Teacher Tapp’s own findings on staff attitudes towards AI reflect the groundswell of differing opinion. 

On one end of the scale, there’s positivity among staff on the issue of AI, expecting it to bring significant change – while others feel AI will dramatically alter education in a damaging way. 

Somewhere in the middle ground, there are those who are positive about its impact but expect slower change, and those who share concerns but believe the effect will be limited.

A workable approach needs to bring different staff perspectives into the same conversation; schools need space to explore what AI should and should not be used for, and how decisions will be reviewed as the technology develops.

The Department for Education has made clear that any use of generative AI by staff, students and pupils should be carefully considered and assessed, with benefits and risks evaluated in context. 

Its guidance reiterates that pupils should only use generative AI in education settings with appropriate safeguards in place, including close supervision and suitable safety, filtering and monitoring features.

This, then, places responsibility on schools to set expectations, allowing leaders to move from reactive responses to a more structured model.

Homework and assessment need a fresh look

Unauthorised AI use challenges how schools check understanding

Discourse on AI misuse is often framed simply as cheating, but the classroom implications go beyond that. 

If a pupil can produce sophisticated outputs instantly, teachers need to know what those outputs prove. Has the pupil understood the subject? Have they evaluated the information? Can they explain their thinking? Can they apply knowledge in a new context? 

Bloom’s Taxonomy is an educational framework that ranks thinking skills from simple to complex. AI has completely disrupted the bottom half of the pyramid. It can easily handle lower-order tasks – like remembering facts and writing basic summaries – in the blink of an eye.

To prove a student actually learned something, however, assignments must now explicitly target higher-order skills: analysing the AI's output, evaluating it for mistakes, and creating original ideas that a computer cannot simply replicate. 

In response, the DfE has advised that schools may wish to review homework policies and other types of unsupervised study to reflect the availability of generative AI. This may include creating guidance for staff and pupils on when the use of AI is acceptable or appropriate.

That could mean designing tasks with clearer expectations. Schools may decide that some assignments should be completed without AI, while others could actively teach pupils how to use AI critically, ethically and transparently. In both cases, the rules need to be visible.

Maintaining integrity in the exam hall

While the focus is often on generative AI in coursework, the integrity of the formal exam hall is under equal pressure. 

Ofqual’s Chief Regulator, Sir Ian Bauckham, recently warned of a growing threat from high-tech smart devices, including smart spectacles and earpieces, which are increasingly difficult to detect. With mobile phone and smart device offences accounting for over 44% of student malpractice cases in the 2025 summer series, the challenge of invigilation is evolving rapidly. 

Schools must now ensure their assessment policies not only address AI-generated content in coursework but also strictly govern the use of wearables and connected devices during formal assessments.

Staff training has to keep pace

Policy only works when people know how to apply it

A written AI policy is a useful starting point, but staff need practical support to translate abstract rules into everyday decisions. Because teachers, support staff, and school leaders interact with AI in different ways, a one-size-fits-all approach simply will not work. 

Each group requires targeted training, distinct boundaries and role-specific examples to navigate the technology effectively.

This makes role-specific Continuous Professional Development (CPD) essential. Schools must build professional confidence in utilising major platforms such as ChatGPT, Google Gemini, and Microsoft Copilot to streamline workloads safely. 

Computeam’s Learning Locker is designed to support flexible, self-directed staff learning across education settings. New AI-focused Learning Locker courses, including content around Google Gemini and Microsoft Copilot, give schools a practical route for developing staff understanding without adding unnecessary pressure to already busy timetables.

We cannot assume everyone starts from the same baseline. Staff readiness ranges from enthusiastic innovators – who use AI daily for planning and admin – to cautious beginners who need help with the absolute basics. Implementing a single, cohesive training pathway gives the whole school a shared playbook; a common language that ensures that school policy, classroom teaching and leadership strategies never work at cross-purposes.

AI leadership needs a whole-school approach

Strategy, policy and CPD should work together

AI governance becomes much more effective when it is connected to a wider digital strategy. Schools and trusts need to consider AI alongside safeguarding, cyber security, data protection, teaching and learning, workload, procurement and digital and technology standards.

Computeam’s Artificial Intelligence Leadership Programme has been developed to support schools and trusts through this process. The programme includes support for building an AI strategy with long-term vision, launching an AI policy, delivering structured staff CPD, introducing an automated AI application approvals process, creating an AI working party, and delivering learner-focused assemblies and workshops.

A tiered structure means schools can choose a level of support that reflects their current position. For some, the priority may be to create an AI roadmap and launch a clear policy. For others, the next step may be a more collaborative model, with an AI working party bringing together leaders, teachers, support staff and technical colleagues.

A collaborative approach is especially useful for multi-academy trusts, where consistency across schools can be a challenge to maintain. AI decisions made in one school may have implications across the trust, particularly when tools process data, affect safeguarding procedures or influence teaching and assessment.

What should schools do next?

A practical response starts with visibility, policy and training

While the reported rise in AI misuse requires urgent attention, it should not trigger panic. 

Schools can respond in a measured, strategic way.

The first step is to understand the current picture. Leaders should review where AI is already being used by staff and pupils, which tools are in circulation, and where the greatest risks sit. This should include homework, assessment, safeguarding, data protection, cyber security and workload.

Computeam supports schools and MATs with AI strategy, staff training, Learning Locker courses, and a new Artificial Intelligence Leadership Programme designed to help education settings move forward with clarity and confidence.

Talk to Computeam about your AI strategy →