SPOKE: Sector Partnership for Optical Knowledge and Education

Artificial intelligence in optometry education: opportunities and challenges

Historically, educational methods in Higher Education (HE) have evolved slowly, but the Covid-19 pandemic and the arrival of ChatGPT in 2022 have led to technological developments that are now driving change at an unprecedented rate. In particular, the latest crop of Generative Pre-trained Transformers (GPTs), such as ChatGPT 4, passed the Turing test in 2023 and so, the once derided ‘chatbots’ are now indistinguishable from humans (Mei et al, 2024). We can’t avoid this rapidly evolving and disruptive technology, which ‘has permeated every facet of the [medical] education environment’ (Gordon et al, 2024), but we do have the power to shape its impact (Rahiman & Kodikal, 2024; Schon et al, 2023). As educators, our challenge is to identify the potential benefits, such as personalised-adaptive learning, enhanced simulation experiences and automated feedback, whilst minimising the risks associated with over-reliance, misinformation and academic misconduct. As a minimum we need to ensure fairness to students, maintain the validity of qualifications, protect data security and maintain public confidence, all key government priorities for the education sector (Ofqual, 2024). Let’s start by looking at some of the opportunities.

For staff and students alike, one of the greatest opportunities, available right now, is the potential to have a personalised, conversational, subject specific expert working with us. Such systems can be created on platforms like ChatGPT, which for paid users, can be used to create Custom GPTs based on the materials you upload rather than a ‘web scrape’. For educators, Custom GPTs could help develop teaching and assessment materials. For students, they could be a multilingual optometry oracle, who never tires of even the most trivial question and is available day or night. Adaptive learning systems improve student grades and satisfaction (Contrino et al, 2024) so this is a positive development. Other benefits available to educators right now include generic tools like Blackboard’s ‘AI Design Assistant’, which can help with course design, Moodle’s integration of contemporary Large Language Model (LLM) tools to help with almost anything, and voice to text systems, like SuperWhisper, which could help alleviate the email burden. The list of potentially helpful education related tools is long and growing. For example, some of the AI tools being used by medical students (Maaß et al, 2024) are:

Many of the challenges we face are shared across the HE sector, but some are particular to optometry. For example, finding enough patients, with a wide enough range of conditions to help our students advance their clinical skills is a perineal challenge, but even here AI systems can help (Cook et al, 2025). For example, platforms such as Sight Test Tutor use LLM technology to bring real patient cases ‘back to life’. Students can talk to virtual patients, record their ‘History and Symptoms’, select the test results they want, formulate a diagnosis and even provide the patient with advice. A second LLM, running in the background compares the student’s performance with a ‘ground truth’ medical record, and via a transparent marking scheme, provides the student with detailed feedback on their performance. Such systems are not a replacement for real clinical work, but they are a useful adjunct helping students build their clinical skills in a patient safe environment.

Whilst these emerging technologies offer an array of opportunities there are challenges too. Academic misconduct is on the rise (Song, 2024), about 10% of students think it’s “okay to cheat if I am not caught” (Lund et al, 2025) and AI plagiarism checkers have been unreliable (Perkins et al, 2023). So, we need to be increasingly careful about how students are assessed. Over reliance on AI systems is another area of concern. A recent systematic review highlights several negative effects associated with over-reliance on AI dialogue systems, such as diminished problem-solving skills, impaired ability to construct logical arguments, reduced effort to craft well-structured sentences, and academic laziness (Zhai et al, 2024). For a more comprehensive discussion about the opportunities and challenges afforded by AI in society more generally, the interested reader might like to consult Dwivedi et al, (2023).

We are living through a time of unprecedented change, machines have become indistinguishable from humans (Mei et al, 2024). Some futurists might welcome a world run by robots, but for most of us that is not the outcome we are looking for. So, we need to harness the power of this remarkable technology to ensure it delivers for our students, and ultimately the patients they care for. Fortunately, it is not the technology itself but rather our attitude toward it, which will ultimately shape its adoption in higher education (Rahiman & Kodikal, 2024). So, we need to be proactive, the revolution is in our hands.