Using generative AI to support revision and learning

Introduction

The use of generative artificial intelligence (GenAI) platforms, such as OpenAI’s ChatGPT, Google’s Gemini and Microsoft’s Copilot, has accelerated in recent years, with 35% of the public using GenAI each month^​1​. AI has many promising applications within pharmacy, potentially transforming how medicines are discovered, dispensed and managed. GenAI is also a powerful support tool for learning.

This article outlines how GenAI can be used appropriately by pharmacy professionals at all stages of their professional development, from students undertaking initial training and education to experienced members of the workforce pursuing credentialling, continuing professional development and lifelong learning.

GenAI is a form of AI which comprises deep-learning models that are able to generate new content, such as text, videos and images, in response to user prompts based on information the model has previously learned^​2​. Tools commonly used in the UK include Gemini, Copilot, Claude and Perplexity, with ChatGPT currently the market leader^​3​. 

In 2026, the General Pharmaceutical Council (GPhC) issued guidance on the use of GenAI for revalidation, as well as a joint statement on its use in health and care professional education^​4,5​. In April 2026, it published its position statement on how pharmacy professionals should approach the use of AI. This statement highlights the roles that AI could play in benefiting patients and the public, such as in diagnosis and prescribing, supporting operational tasks and as a tool to help when writing revalidation submissions^​6​. One of the benefits of using GenAI is its ability to summarise information and produce a useful output in seconds. Below, we discuss the importance of being specific with prompts, examine considerations when using GenAI, and highlight a selection of ways it can be used to support revision and learning.

Producing specific prompts

In terms of AI, a prompt is an instruction, question or request given to an AI model to guide its output. The words used to prompt GenAI will affect the quality of the response it produces. The more specific the instruction, the more likely the output will match what the user wants. For example, in Figure 1, the image on the left was produced by Copilot^​7​, using the prompt: “Produce a cartoon-style image of someone revising using AI.” The image produced was appropriate to the prompt; however, we wanted the person to resemble an adult learner. By amending the prompt to state “an adult” rather than “someone”, the image on the right was produced. 

Figure 1: Images of a person revising with help from GenAI, demonstrating the importance of being specific with prompts

Important considerations when using GenAI

Content produced by GenAI is only as reliable as the information it has learned. When using GenAI, it is important to adopt a critical approach and treat outputs as useful starting points rather than indisputable facts. AI models have been known to give biased responses that might include fake references, which are known as “hallucinations”, in its output. These hallucinations occur because GenAI creates plausible text from patterns in its training data rather than retrieving and verifying real sources^​8​. These systems are often designed to produce helpful and agreeable responses. When using GenAI, it is therefore vital to review the output against verified source material to ensure the information it has provided is correct. 

Another consideration surrounds data protection and management. GenAI learns from the information put into it, which includes the prompts that users enter and documents that might be uploaded to it. It can then relay this information to other users who use similar prompts. Content should not be uploaded without the author or copyright holder’s permission. This includes things such as unpublished teaching materials from an instructor or education provider, as well as pages from books, articles, videos or other materials where there could be a breach of copyright or other ownership rights. Confidential information (e.g. data relating to patients) should never be used. Some GenAI platforms offer a ‘closed’ or ‘private’ system, where data stays within the host organisation, offering an additional layer of data protection. This is why some organisations are specific about the platform that needs to be chosen when using GenAI and why GenAI use is prohibited in certain circumstances.

Using GenAI in revision and learning

GenAI’s ability to rapidly process large volumes of information provides many potential benefits when revising for assessments. Learners can use GenAI to produce summaries of lengthy documents and rephrase complex information into simpler terms, reducing the time needed to produce materials such as revision notes and flash cards. GenAI can also be used to produce mock exam questions to practice with, which is something learners are known to have a preference for as part of revision^​9​. These materials could then be used in combination with more traditional content available to learners through other channels such as GPhC registration exam practice papers or ONTrack.

The examples below illustrate how GenAI can be embraced to support learning and include some suggested prompts.

Summarising information

When learning to be, and practising as, a pharmacy professional, it is common to encounter lengthy, complex documents such as clinical guidelines, formularies and research papers. GenAI can summarise this information so that it is easier to digest. A consideration when using GenAI in this way is that the interpretation of the information by the chatbot and learner may be different. GenAI is therefore best used as a tool to help develop an initial understanding of the content prior to reading it fully and should not be used as a shortcut. By reviewing the content with a better understanding, you will also be able to ensure that the GenAI’s summary is accurate.

An example of using GenAI in this way is to ask the chatbot to summarise the initial steps of pharmacological treatment of hypertension based on current National Institute for Health and Care Excellence (NICE) guidance.

Suggested prompt: “Summarise steps one and two of the pharmacological treatment recommendations of NICE NG136.”

Using GenAI to produce revision materials

Some learners prefer to combine their lecture notes with additional information that they have read, create flash cards to support knowledge recall or produce spider diagrams to organise complex information. These are outputs that GenAI is capable of generating, potentially saving a lot of time. For example, lecture notes could be attached and the chatbot prompted to provide additional information around a specific topic from the notes or to make suggestions for further reading. However, the downside of this is potentially missing out on the benefits that come from developing revision materials manually. The act of reading, synthesising the information and then summarising it to produce revision materials forms the basis of knowledge retention and promotes deeper engagement and understanding.

Learners are also known to prefer one or a combination of visual, auditory, reading or kinaesthetic learning approaches^​4​. This gives rise to another potential use of GenAI as it could be used to reformat revision materials from one learning format into another. For example, GenAI could be prompted to produce a series of flash cards for visual learners or to produce an audio deep dive into a specific topic for auditory learners. 

If you are unsure of your preferred learning style, the VARK learning questionnaire is linked in the ‘Useful resources’ section below.

Suggested prompt: “Produce a set of flash cards for revision on how to perform a respiratory examination on a patient.”

Using GenAI to produce practice exam questions

Preparing for exams using practice or sample questions is known to be a highly effective learning and revision strategy^​9​. However, practice questions are not always made available by the education provider, or the number of examples might not match what the learner would like. In this situation, learners might attempt to write their own practice questions, use older questions they find online or questions originating from outside the UK. This runs the risk of working through questions that might not accurately reflect the learning outcomes of the assessment or fail to align with the exam format. An alternative is to prompt GenAI to write practice questions on a particular topic, specifying the question format and including specific details such as the number of distractors (i.e. incorrect answer options) for multiple choice questions. 

Suggested prompt: “Produce a 10-question mock examination on cardiovascular pharmacology aimed at Year 3 pharmacy students. Use the multiple-choice format with 4 distractors.”

If there is a particular area to focus on, such as dilution calculations, the prompt should include an instruction to specifically provide questions on this theme.

Suggested prompt: “Produce a ten-question pharmaceutical calculations paper based on the GPhC’s pharmacy registration exam format. Include examples of dilution and infusion rate calculations.”

If using a closed / private system, an additional approach could be to attach the lecture notes to the prompt and ask the AI tool to produce questions based on the attached document. As mentioned previously, it is important to respect ownership of these documents. They should not be uploaded to ‘open’ GenAI systems, as this may result in the content being retained and later generated in responses to other users. 

Helping with extended writing

When tasked with writing an essay, a reflection or other assignment, it is sometimes difficult to determine the best way to structure the piece. This is another area where GenAI can be useful — not to write the actual document to be submitted but instead to make suggestions on things such as structure and possible ways to improve spelling, grammar and the brevity of the writing. When using GenAI in this way, it is important to take care not to breach the rules and guidance issued by the organisation or education provider. Such use aligns with the GPhC’s guidance, which makes clear that GenAI must not be used to write any revalidation submission or to falsify information to be included in the submission, but that its use to formulate ideas and correct grammar is permitted if GenAI is used in a responsible, transparent manner^​10​. Transparency on GenAI use is important in taking accountability when using GenAI and is discussed further in the ‘Best practice and transparency’ section below.

Suggested prompt: “Suggest how to structure the background section to an assignment on pharmacogenomics.”

Role playing objective structured clinical examinations and other interactions

A creative way that GenAI can be used as an educational tool is through the role play of patient, carer or colleague interactions. This can be particularly useful when preparing for objective structured clinical examinations (OSCEs). Some platforms offer spoken responses to the learner’s questions or information to provide realistic practice. These interactions may be based on pre-designed scenarios (as seen in platforms such as Geeky Medics) or on custom scenarios created by the learner (as seen in platforms such as MLA Buddy). While these platforms often provide basic access for free, full functionality often requires a subscription, which limits accessibility. However, with appropriate prompting, it is possible to role play with more widely available GenAI chatbots. For example, a chatbot could be prompted to act as a patient by providing a clear scenario, such as: “You are a patient newly diagnosed with moderate asthma, and I am a pharmacist whose task is to counsel you on how to use your inhaler.” 

It is then possible to practise asking questions, taking a history or providing counselling within a simulated interaction. To better reflect OSCE conditions, voice mode could be enabled, strict time limits imposed and prompts could be refined to increase scenario complexity or target specific skills. Although GenAI cannot replace human-simulated patients or peer-to-peer role play — as GenAI lacks important aspects of communication such as tone and body language — it has the advantage of relative unpredictability in its responses. This requires users to respond to cues and adapt to new information, reflecting the challenges of real-life patient and OSCE station encounters. After completing a role play, GenAI can be prompted to assess performance and provide immediate feedback based on a typical OSCE assessment checklist. Alternatively, a scenario could be inputted and the chatbot asked to generate an OSCE-style checklist to act as a guide. 

Best practice and transparency

As mentioned above, the GPhC is one of the healthcare regulators to have published guidance on use of GenAI, with other organisations having their own codes of practice in place. It is therefore essential to review these guidelines before starting to use GenAI as a learning tool to ensure information is being processed in a safe, responsible manner. If using GenAI, a transparent approach should be taken to ensure academic integrity standards are maintained. For some organisations, this may require the completion of AI training and the inclusion of a disclaimer stating when GenAI has been use — for example, as part of an assessment submission. The most important message above all is that if advised specifically not to use AI, it must not be used, even for tasks such as summarising or making suggestions on structure.

An example of the GenAI usage statements used at University of Birmingham^​11​ would be:

• “No content generated by AI tools has been presented as my own work;
• I acknowledge the use of [AI system(s) and link] to generate materials for background research and self-study in the drafting of this assessment;
• I acknowledge the use of [AI system(s) and link] to generate materials that were included within my final assessment in modified form. The following were initially generated by AI…”

Summary

GenAI has many potential uses within pharmacy. Here, we discussed a small number of ways it can be used to support revision and to help users engage with, and better understand, complex information. Use of GenAI is developing rapidly and becoming more widely accepted across society and within healthcare; however, there will be times where GenAI use would be inappropriate or should be avoided.

By taking a critical approach to evaluate the accuracy of information provided and utilising GenAI in a way that suits personal learning preferences, it can become a powerful, efficient tool that supports preparation for assessments and other aspects of lifelong learning.