AI vs AI: The conflict between Artificial Intelligence and Assessment Integrity

Artificial Intelligence (AI) tools have seen a substantial increase in use, now integrated with many aspects of our day-to-day lives.

• 45% of large businesses in the UK are now using at least one type of AI technology, an increase of 13 percentage points in a year. source

• A report published in January this year reported “73% of the public have used AI in their day-to-day life in the past month.” source

These statistics are significant, both in terms of their scale and also the impact on how people work – both in their personal and professional lives.

We have likely all seen some of the benefits AI can provide, completing tasks in seconds which would otherwise take days. Not to mention the potential implications for healthcare and science.

Where we must however proceed with caution, is when it comes to credibility of work. Recognising AI as a tool which supports the human-led production of a piece of work, rather than a tool which completes the work itself, is key.

In the assessment world, this is particularly pertinent.

Ofqual’s Chief Regulator, Sir Ian Bauckham, wrote to Awarding Organisation leaders on the 2nd March 2026 regarding the misuse of AI in assessments. In the letter he says:

“Where coursework exists, it is intended to contribute to the learning of students as well as being a method of assessment. The use of AI in coursework deprives students of that intended learning experience, because the work is done by the AI rather than by the student. If use of AI is undisclosed, then the integrity of the qualification is damaged as marks no longer fully reflect the student's own work.”

The full letter can be read here: Letter to awarding organisation chief executives about malpractice - GOV.UK

Accountable to Ofqual, Assessment Organisations must treat the submission of AI-generated work as plagiarism, when it is presented as original work. Assessment Integrity is central to the reputation of, not only the Assessment Organisation in question, but also the wider Education and Assessment ecosystem.

Assessments do not solely exist to validate knowledge. They are designed to help learners understand concepts, think critically, and develop skills. By relying too heavily on AI when producing work for assessment, learners risk bypassing the process of research and analysis. Knowledge becomes superficial rather than meaningful. Learning happens through effort, mistakes, and reflection.

Critical thinking is one of the most valuable skills education builds. Producing reports or projects helps learners to evaluate evidence, form arguments, and develop their own perspectives. Over time, relying on AI can reduce confidence in forming ideas and arguments.

While AI should not replace learner effort, it can be useful when used responsibly, to support the work being produced rather than producing it entirely.

For example, during the research and preparation stage of producing material AI can support with brainstorming ideas or used to give a clearer explanation of a difficult concept.

All Microsoft applications provide spelling and grammar checks. We would discourage the use of any other AI tool for this purpose as this can show as a document being AI generated when checked.

If a learner uses AI, they should cite this in their work. When citing AI, learners should always include:

• The tool name (e.g. ChatGPT, Copilot, Grammarly)

• The version or model (if known)

• The date of use

The most common citation style used in the UK is Harvard style, but this is not mandatory and others can be accepted.

Learners should also declare how AI was used, so for example:

“I used ChatGPT to help brainstorm ideas and improve the clarity of my thinking. All final content and analysis are my own.”

Learners should be encouraged to keep a record of the prompts they used and any of the outputs they relied on to help them in producing their assessment material. Learners do need to remember that citation does not make inappropriate use of AI acceptable, for example a complete re-write of a document by AI which is cited as such, would not be accepted.

AI is a powerful tool, but it should not take the place of genuine original work. Assessments exist to help learners grow intellectually, develop critical skills, and build confidence in their abilities.

When used correctly, AI can support learning without compromising Assessment Integrity.

For more information about acceptable use of AI in assessments, please review our Plagiarism Policy. If you have any questions at all please contact our team using the form below.