[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-620908-105":3,"doc-detail-620908-en":81,"detail-sidebar-cat-0-en-105":98},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","partnering-with-ai-to-enhance-fair-and-effective-assessment","Partnering with AI to enhance fair and effective assessment","","AI is increasingly integrated into education, extending from personalised learning to administrative work, and now into assessment. AQA’s Research and Innovation team investigates how AI can complement expert judgement to improve assessment quality, efficiency, and fairness while keeping ethics and human oversight central. The document outlines an ethical AI framework, focuses on explainability for marking transparency, addresses accuracy challenges in handwritten responses, and evaluates AI-generated feedback in classroom trials.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/partnering-with-ai-to-enhance-fair-and-effective-assessment/620908/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/partnering-with-ai-to-enhance-fair-and-effective-assessment/620908.png","ImageObject",300,407,{"name":42,"@type":43},"Jordan Avery","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-30","2026-10-20",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",7,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"How does AQA plan to ensure AI supports fair assessment without losing human oversight?","Question",{"text":63,"@type":64},"AQA uses an ethical AI framework to guide responsible use, with risk assessment and governance in place. The approach keeps ethics and human oversight central while using AI to complement expert judgement.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"What does AQA say about making AI decisions in marking more transparent?",{"text":68,"@type":64},"Explainability is a key goal, including analysing how AI models focus on parts of a student’s answer. Findings suggest models can highlight influential words and phrases even without a formal mark scheme.",{"name":70,"@type":61,"acceptedAnswer":71},"How is handwriting digitisation addressed before applying AI to exam responses?",{"text":72,"@type":64},"Because most UK exams remain handwritten, responses must be digitised first. AQA piloted a method using subject-specific dictionaries and AI models to reduce transcription errors, improving accuracy without needing human transcribers or bespoke handwriting software.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},620908,1792901253,{"code":4,"msg":82,"data":83},"success",{"doc_id":79,"user_id":84,"nickname":42,"user_avatar":85,"doc_module":4,"category_id":86,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":87,"file_id":88,"file_url":89,"file_type":90,"file_size":91,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":26,"language":92,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":93,"faqs":94,"seo_title":95,"seo_description":12,"update_tm":96,"read_time":97},1099523882367,"https://ap-avatar.wpscdn.com/davatar_9964176cb1d06d4a9deccf72a44ae3dc",8,"Partnering with AI to enhance fair and effective assessment  \nAI is becoming a familiar presence in education, supporting everything from personalised learning to administrative tasks.  \nAs its role expands into assessment, AQA’s Research and Innovation team is exploring how AI can complement expert judgement – improving quality, efficiency and fairness, while ensuring that ethics and human oversight remain central.  \nA framework for ethical AI  \nAt AQA, an ethical AI framework – developed in collaboration with external consultants – guides the responsible use of AI across the organisation. This framework underpins risk assessment and governance as we explore the potential of AI tools to enhance assessment practices.  \nMaking AI decisions more transparent  \nExplainability is a key goal across AIresearch – especially when it comes to marking, where transparency and fairness are critical. At AQA, we’ve explored this by analysing how AI models ‘pay attention’ to different parts of a student’s answer. Early findings suggest these models can highlight the words and phrases that influence their decisions, even without a formal mark scheme. This could help build tools that support human judgement and make AI decisions more transparent.  \n[aqa.org. uk/about-us/our-research](aqa.org. uk/about-us/our-research)[ ](aqa.org. uk/about-us/our-research)[research@aqa.org. uk](research@aqa.org. uk)  \n @AQA  \nImproving AI accuracy for handwritten responses  \nSince most UK exams are still handwritten, responses must be digitised before AI tools can be applied. But handwriting recognition  \noften introduces errors. To address this, we piloted a method that uses subjectspecific dictionaries and AI models to cleanup transcription errors. This significantly improved accuracy, enabling more reliable use of AI without the  \nneed for human transcribers orbespoke handwriting software.  \nExploring AI-generated feedback in the classroom  \nAs part of a collaborative research project with a university partner and a secondary school, we explored how AI-generated feedback could support learning in the classroom. We trialled AI-generated feedback on short English essays, working with both students and teachers. The study showed that AI can deliver rapid, detailed and personalised feedback, while also highlighting the importance of aligning AI use with appropriate question types.  \nTeachers valued this AI support but stressed the need for thoughtful implementation topreserve the student–teacher relationship.  \nSupporting examiners through AI innovation  \nAQA is proud to be leading the only education-focused project within the latest round of the UK government’s £97 million Prosperity Partnership initiative – a major investment aimed at advancing AI innovation through collaboration between academia and industry.  \nIn collaboration with King’s College London, we’re developing a safe and effective virtual AI assistant to support examiners in marking GCSE and A-level exams. This tool aims to enhance fairness and consistency by providing an additional layer of quality control. It could also have wider applications in other expert-led fields where accuracy and trust are critical.  \nLooking ahead  \nAI will play an increasingly important role in assessment – not by replacing human expertise, but by enabling and enhancing it. At AQA, we’re committed to integrating AI in ways that empower professionals, uphold fairness and build trust across the education system.  \n[aqa.org.uk/about-us/our-research](aqa.org.uk/about-us/our-research)[ ](aqa.org.uk/about-us/our-research)[research@aqa.org.uk](research@aqa.org.uk)  \n @AQA  \nCopyright © 2025 AQA and its licensors. All rights reserved. G02762-02","cbCaissJKuIpynsE","https://ap.wps.com/l/cbCaissJKuIpynsE","pdf",292183,"English","# Partnering with AI for fair assessment\n## Ethical AI framework and governance\n## Making AI marking decisions more transparent\n## Improving accuracy for handwritten responses\n## Exploring AI-generated feedback in classrooms\n## Supporting examiners with a virtual AI assistant\n## Looking ahead","[{\"question\":\"How does AQA plan to ensure AI supports fair assessment without losing human oversight?\",\"answer\":\"AQA uses an ethical AI framework to guide responsible use, with risk assessment and governance in place. The approach keeps ethics and human oversight central while using AI to complement expert judgement.\"},{\"question\":\"What does AQA say about making AI decisions in marking more transparent?\",\"answer\":\"Explainability is a key goal, including analysing how AI models focus on parts of a student’s answer. Findings suggest models can highlight influential words and phrases even without a formal mark scheme.\"},{\"question\":\"How is handwriting digitisation addressed before applying AI to exam responses?\",\"answer\":\"Because most UK exams remain handwritten, responses must be digitised first. AQA piloted a method using subject-specific dictionaries and AI models to reduce transcription errors, improving accuracy without needing human transcribers or bespoke handwriting software.\"}]","Partnering with AI to enhance fair and effective assessment | PDF",1792464738,5,{"code":4,"msg":82,"data":99},[100,104,108,112,116,121,125,128,133,136,140],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":101,"show_sort_weight":102,"slug":103},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":105,"show_sort_weight":106,"slug":107},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":109,"show_sort_weight":110,"slug":111},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":113,"show_sort_weight":114,"slug":115},"Comic",60,"comic",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":118,"show_sort_weight":119,"slug":120},6,"Technology",50,"technology",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":122,"show_sort_weight":123,"slug":124},"Healthcare",40,"healthcare",{"id":86,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":126,"slug":127},30,"research-report",{"id":129,"doc_module":4,"doc_module_name":25,"category_name":130,"show_sort_weight":131,"slug":132},9,"Religion & Spirituality",20,"religion-spirituality",{"id":131,"doc_module":4,"doc_module_name":25,"category_name":134,"show_sort_weight":131,"slug":135},"World Cup","world-cup",{"id":137,"doc_module":4,"doc_module_name":25,"category_name":138,"show_sort_weight":137,"slug":139},10,"Lifestyle","lifestyle",{"id":141,"doc_module":4,"doc_module_name":25,"category_name":142,"show_sort_weight":97,"slug":143},19,"General","general"]