There is a need for the UK education system to recognise the rapidly changing world and increasing importance of preparing students for the future in terms of careers and the challenges in part related to technology. Many current people in the workforce already have a fear of losing their jobs to Artificial Intelligence (AI)1 and are worried about their skill level in understanding the changing face of technology and its seemingly infinite and complex ways of working.
The IB has a broader strategy to embrace digital learning for both students and educators; an active choice to see the opportunities and benefits and keep pace with the potential future2. They emphasise a curriculum that works on both specialised knowledge and the "soft skills" like teamwork, communication, and problem-solving that are increasingly valued in the modern workforce.
At WA, a year ago, we discussed what this would look like for our students and educators in our context and came up with the following ideas:
The ambition we have is to create students who are highly effective at using different digital tools and platforms to support their learning, critically evaluating information sources, communicating their work, solving problems, designing inquiries and, in doing so, show increasing autonomy in their learning process. Students can communicate and collaborate effectively online and understand digital ethics, reflecting regularly on their use of technology.
We would like to support educators to be highly developed in the use of technology whereby they integrate digital tools into their instructional practices. Evidence of not just using technology but showing how it can enhance learning to others as a pioneer.
This was the focus of one the Professional Learning Communities (PLC) for this academic year and I’m delighted to signpost some of their learning more formally in this blog post.
--------------------------------------------------------------------------------------------------------------------------------
Ben Law Smith’s inquiry was “Can GenAI be used as a tutor to improve attainment in Year 10 GCSE Citizenship Studies?”
Recent studies suggest AI tutoring can have the same impact as in person tutoring on 11-18 year olds (Lie et al, 2025) and that it can improve test results by an average of 20% after 9 months use (Swarglary, 2024).
Ben did a 5 month quasi-experiment with Year 10 Citizenship students to see if he could recreate these findings at WA.
One class (out of three) was introduced to a bespoke "'CitizenshipGPT' which was pre-trained to act like a tutor and use the AQA exam specification materials instead of general information. Students could only use it outside of class.
He carried out Pre and Post experiment surveys to understand usage and monitored attainment over four assessments including an end of year GCSE mock exam.
Findings:
Students are interested in using GenAI to learn - before the study, 69% of people who had used GenAI at least once before wanted to use it at school to learn.
Access to a GenAI tutor makes students more likely to be interested - after the study, 86% of students who used the GenAI tutor would like them to be used in other classes.
Frequent use maintained On Track grades - students who used the GenAI tutor 'Many times' in the access period (3 students) maintained their On Track grades from Semester 1 to Semester 2. By comparison, the ON Track grades for students who used it 'A few times', 'once or twice' or 'never' on average reduced by -0.4.
Challenges:
Adoption - Only 58% of sample tried CitizenshipGPT and of those that did, 79% used it ‘Once or twice’ or ‘A few times’, not enough to see impact. As a result the sample size was very small and findings may be unreliable.
Free version vs. paid version - the free version of ChatGPT that most students use is a much less powerful model so the explanations and responsiveness to their level may not be as good
Whilst not conclusive, this short experiment has shown that GenAI is something students are interested in using and appears to support learning as a revision tool.
However, correct set up and use is critical - research shows that using GenAI as a ‘socratic tutor’ can have a positive impact on attainment whilst using it as an ‘answer machine’ can harm attainment. This is most memorable way I’ve seen this idea put: ‘If you went to a gym and got a robot to lift weights would you get stronger? Obviously not. However, if you got a robot to motivate you to go to the gym, then encourage and guide you to do exercise in the correct way week after week, would you get stronger, faster? Yes! Don’t use GenAI to do your heavy lifting for you!’
"Many schools across the world (especially in Australia under the auspices of state government) are finding their own trials of GenAI tutors have benefited their secondary students - I hope to continue looking into whether it could be a benefit to ours too!"
--------------------------------------------------------------------------------------------------------------------------------
Andrew Compton’s Inquiry was “How can AI-Generated Imagery support the presentation of assignment work in BTEC National Sport Studies?”
Andy worked with Year 12 and 13 BTEC Sport students to examine the creation of AI generated imagery and its use in providing suitable accompanying images to enhance coursework.
Over time, they discussed and experimented with different typed statements and adapting ways in which they make requests of the AI model, with the aim of improving the quality and usefulness of the images generated.
Every single image below has been generated using AI technology, using either Gemini 2.5 flash, or ChatGPT 4o mini (free version). They are a sample of some of these images along with a short commentary on the successfulness of their creation and application of their use as part of the BTEC Sport course.
Injuries in Rugby
Prompt: ‘Create me an image to show a rugby player suffering injury due to a lack of flexibility’
Students had to discuss intrinsic risk factors for injury in Rugby Union. One of the available options was inadequate flexibility and how it can lead to injury.
The student wanted to produce an image to supplement their text response explaining how this can lead to injury and what we can do to prevent such injury.
The AI image create has focused on showing a muscular rugby player grimacing, while unable to execute a stretch effectively.
The image contributed to a very comical discussion in the lesson on its suitability to be used for the assignment, but we agreed that it was able to get across the point that the student wanted to make, and depict the injury risk in question.
Skeletal System
As part of their research and revision for their Anatomy and Physiology unit, BTEC Sport students tried to create diagrams of different body systems.
One student was able to use AI to create a fairly useful diagram of the different sections of the spine, but it lacked the detail to differentiate between different individual vertebrae.
When they tried to specify, the AI was unable to label the sections effectively and could not number each of the individual vertebra effectively. Spelling errors and some incidences of repetition were an issue in both images.
Rondo Passing Drill
A student wanted to visually represent the ‘rondo’ passing drill that they were planning to include as part of their coaching session. Below are three of the incarnations that AI created following refining of the request. None of the images could quite achieve what the student wanted to see.
The first image has two blue shorts in the outer circle and the ball at the feet of the defender.
The second image fixes the shirts but there are now three balls in play and a confusing circular arrow
The third image pits 16 attackers up against a lone defender and they all have a ball each!
A number of other attempts were made, but unfortunately the AI could not produce an accurate reproduction of the rondo drill being set-up or executed correctly.
Basketball Hierarchy
Prompt - ‘can you create me an image to show the hierarchy pyramid of basketball organisations from FIBA to grassroots’
Students were asked to provide a visual representation of the hierarchy of their chosen sport to demonstrate how rules and regulations are passed down from rule makers to grassroots competitors.
This effort is initially visually appealing and gives a relatively overall accurate reflection of the different tier levels. However, on closer inspection we can see a few spelling errors and some incorrect or obscure titles on the pyramid tiers.
This did not turn out to be an issue, as the image was only used to accompany a detailed description of how the hierarchy works, but was quite impressive for a first time effort, and with some refining, could maybe develop into an excellent image.
Overall, the results of this trial were varied, and many images were more ‘style over substance’ than truly effective at enhancing the presentation of the work. Through trialling and refining the requests, they were able to put success down to being dependent on a number of key factors:
The range and availability of images on the web that the AI generator needed as input
The power of the AI generator used
The detail and breadth of information given to the AI generator through each request
The complexity of the request made of the AI generator
The ability of the image requester to build upon the request/question allowing the AI generator to refine the image
--------------------------------------------------------------------------------------------------------------------------------
Using the guidance produced by Ben Law Smith, David Madden generated Chat bots to aid his Year 13 AISL and Year 11 Higher GCSE students with their Maths revision ahead of their exams. He subsequently refined the prompts that he gave the bots (by asking ChatGPT to refine them for him!). The support that the bots provided seems to be more personalised as a result of the refinement. It started by asking them what they knew about the topic; it then modelled the maths and then gave the student an question to work through - this bit is set up in such a way that the student can type in their working and then at the end they are told to say this is the final answer so they submit that and that gets marked.
In parallel to this, Tom Pattinson, from the We Are in beta community reached out to ask David a question about the use of AI in the classroom: “I shared my links with him and I am now part of a new AI community hosted by the community! Through this community I have gained access to the recently produced DfE guidance on the use of Generative AI in the classroom: Generative artificial intelligence (AI) in education
This is certainly work in progress but it will be an exciting journey! More to follow!”
--------------------------------------------------------------------------------------------------------------------------------
Elizabeth Anthony’s Year 11 science class was a mixed ability nurture group with varying levels of needs. She wanted to explore the use of AI to help them access learning better and build their confidence. After teaching them about magnetism, Elozabeth wanted to see if she could use AI to create a comic strip based on the scientific information they need to know, but tailored to their level and needs. After several trials, she was able to produce one which she then used to deliver an activity:
Activities that can be done using this strip:
Comprehension task - read and answer Qs.
Find misinformation and correction on the strip.
Use this to write down key facts they have learnt from the strip.
What was the impact?
Students can work at their own pace.
Comprehension - helps improve literacy skills.
Dual coding - if you choose the right image.
Students are more likely to be engaged and not see that they are learning.
Another exploration of AI for Elizabeth came after her class found it difficult to access exam questions - low scores were diminishing their confidence. “I used AI to make exam questions more accessible to them. Once they saw that they could do these questions, it made them try better, and they did better in their mock exam, which used the same exam papers as everyone else.”
Adapted version:
AI can be used effectively to help students engage better by providing activities that are tailored to them. This can be used to provide independent work for students on either end of a mixed-ability class, while the middle gets on with their work. Elizabeth was better able to adapt to the needs of her class with the support of AI!
--------------------------------------------------------------------------------------------------------------------------------
We cannot ignore the impact AI is having on our brains3 however it is clear that there is a place for using AI to support learning when used in a way that still asks the students and educators to think deeply and think critically. The responsible use of technology is an ongoing debate and whilst we find the balance it’s exciting to know we can elevate our lesson planning and support for students to progress with tools that also save us some time - time that can then be redistributed back into building meaningful relationships and pastoral care for our students.
What are your thoughts and experiences with using AI?
How are you balancing the use of AI with still being able to think about the task?