Can GenAI be a force for good for an inclusive student experience, or is it yet another tool that can drive inequality and increase attainment gaps?

As a Theme Lead for Inclusive Student Experience at the Bioscience Scholarship Centre (Inclusive Student Experience | Nottingham Trent University), I am always looking at how our scholarship and teaching can be made accessible for all.

Is GenAI inclusive?

To answer the question, we need to understand what GenAI is first. Perhaps the simplest definition is that GenAI is a set of tools that can generate or transform text, images, code, and data.

This post argues that GenAI can support an inclusive student experience when we provide equitable access and give students explicit guidance on its use. We should also teach students to document their GenAI use reflectively in their learning, to preserve their original voice.

GenAI can help students develop new skills?

If AI is going to be part of career readiness, students coming out of university with AI literacy will have an advantage. At NTU, Biosciences courses have Professional Skills modules, and this is an ideal home for these skills to be taught.

AI literacy is not just being able to use an AI prompt box. It is about knowing which keywords to use and how much detail to include in the prompt. This is sometimes called prompt engineering.

In the example below, you can see a simple prompt versus a more detailed one:

Figure 1. On the left, a painting of a flower using Microsoft Copilot (GPT-5.2). Generated on 26 January 2026. On the right, an image of a flower generated from a complex and specific prompt. Generated using Bing Image Creator (DALL-E 3) on 26 January 2026.

The simple prompt often produces generic outputs and might not actually meet your coursework requirements! Explaining to students how, and why, the prompt is so important will supercharge their skills.

Critique and fact-checking

Prompting for critique of text is just as important: students need to know how to verify AI-generated text and find primary sources to back up its claims. They need to be able to spot bias in answers and understand GenAI limits, including AI hallucinations and straight-up mistakes.

This is fact-checking adapted to our times.

GenAI has the advantage of being trained on vast amounts of data, far more than any one human can read alone. If you have ever looked for a journal paper that would help reinforce your findings, you will understand the power of searching large databases! Sources still need to be verified and read, but now they can be found more easily.

This can particularly remove barriers to participation when access to the AI tool is provided by the institution: it can level the field in our diverse student cohorts’ digital literacy and reduce the effects of digital poverty.

On the flip side, some recent studies suggest that using AI without caution or a plan can lead to loss of skills, as the brain does not get to activate regions important in learning. This is particularly important in critical thinking, writing, and skills development.

To add to this, there are issues in terms of accuracy, ethical considerations, logical fallacies, and hallucinations. Used without critical thinking, GenAI can bring “wrongness at scale”.

GenAI can generate what looks like a probable answer very confidently. It does not necessarily mean it is correct. Students learning about a subject will not necessarily know that the data or a theory is wrong. This has caught out consulting companies and even lawyers citing fake case law!

GenAI responses also depend on the data it was trained on. There is a lack of transparency in what data was used to train these models. Any bias that might exist in those data could therefore mean that GenAI responses perpetuate biases and prejudices.

Are non-American and non-European sources considered? Are all sides of the coin considered in the answers our students might be reading?

Accessibility

Once students have found sources and considered GenAI outputs, we can give them the skills to help them synthesise information into their own work and not just pass AI text off as their own.

This is where the skills from Professional modules come in handy. We have a unique opportunity to teach students how to make their content accessible from the outset. We can ensure that the GenAI outputs they generate are usable by screen readers, for example. They can ask outputs to include alt text with images or provide different modes of expression of the same idea, such as a text summary and a diagram together.

For students with disabilities, using GenAI can help them overcome hurdles in writing and give them another tool to participate. Often AI can take spoken inputs and correct typos, which can benefit students with dyslexia, for example, or students with disabilities affecting their hands.

Expressing thoughts

GenAI could support inclusivity by helping all students, particularly multilingual students, express their thoughts better. It could act as a ‘super-version’ of a spelling and grammar checker.

To help students learn how to express their thoughts in an academic manner, GenAI can scaffold their grammar and structure. Over time, in combination with academic writing courses and library webinars, they may reduce their dependence on their GenAI ‘tutor’ as they develop their own academic voice.

The biggest negative that I can see here is the bulldozing of identity or the loss of diverse voices: a lot of the work that comes out of AI sounds the same!
I have seen this in student emails and work they have handed in. These outputs homogenise diversity in style and thought. Student vocabulary and unique expressions are lost, even within the limits of academic writing.

In fact, one of the informal ways proposed to detect GenAI writing is the number of these so-called ‘dead words’, such as “augment”, “delves”, “meticulously”, and even “in today’s rapidly evolving market”.
Reading this writing stops being a journey of discovery and learning and ends up being something of a déjà vu!

Conclusion

The fact is, using GenAI at university is complicated. I believe GenAI can reduce barriers and widen participation, especially for students who may be experiencing university without knowledge of hidden assumptions within the curriculum.

However, we need to take an active role in training our students so that we do not end up with ‘robot voices’ taking over our students’ own thinking.

If we fail in this, we risk losing our students’ trust and will end up widening attainment gaps, as only some students will benefit from this tool.