A
- Academic Integrity
- The commitment to honesty, trust, fairness, respect, responsibility, and courage in all academic work. In the context of GenAI, academic integrity extends to transparent and ethical use of AI tools. Read more on the Principles page.
- Agentic AI
- AI systems that can plan, act, iterate, and make decisions across multiple steps with limited human input. Unlike earlier generative tools, agentic AI can autonomously break tasks into sub-goals, retrieve and synthesise information, and execute workflows. Read more on the Agentic AI page.
- AI Detection Tools
- Software designed to identify AI-generated content. Current evidence indicates these tools produce both false positives and false negatives and cannot reliably determine authorship or intent. Read more on the Risk & Detection page.
- AI Integration Scale
- A framework developed by Leon Furze outlining five levels of GenAI integration in assessment: No AI, AI Planning, AI Collaboration, Full AI, and AI Exploration. Read more on the AI Integration page.
- AI Literacy
- The capacity to engage with AI critically, ethically, and effectively, including the ability to evaluate AI outputs, understand appropriate use, and recognise the role of human judgement. Read more on the Principles page.
- AI-Integrated Assessment
- See Track 2.
- AI-Restricted Assessment
- See Track 1.
- Authentic Assessment
- Assessment tasks that reflect real-world applications and require students to apply knowledge in realistic or discipline-relevant contexts, rather than reproducing information. Read more on the Framework page.
B
- Bloom's Taxonomy
- A hierarchical model of cognitive skills traditionally used in educational assessment. Phillippa Hardman has proposed alternatives focusing on skills AI enhances rather than replaces. Read more on the Framework page.
C
- Cognitive Offloading
- The risk of students delegating thinking, planning, and decision-making to AI systems rather than engaging meaningfully with the learning process. Particularly relevant with agentic AI. Read more on the Agentic AI page.
- Completion-Based Assessment
- Early-stage assessment components graded on a pass/fail or checklist basis rather than full criteria, used to evidence learning progress without significantly increasing marking workload. Read more on the Strategies page.
D
- Discursive Control
- An approach to academic integrity relying on rules, warnings, and AI bans. Contrasted with structural redesign, which changes the mechanics of assessment itself. Read more on the AI Integration page.
E
- EU AI Act
- European legislation that highlights education as a high-impact context requiring transparency and risk mitigation in the use of AI systems. Read more on the Framework page.
F
- Formative Assessment
- Assessment designed primarily to support learning and provide feedback, rather than to assign grades. Process-based approaches often include formative elements. Read more on the Strategies page.
G
- GenAI (Generative Artificial Intelligence)
- AI technology capable of generating text, images, code, and other content. Since the launch of ChatGPT in November 2022, GenAI has significantly impacted higher education assessment. Read more on the Framework page.
- Graduate Attributes
- The qualities, skills, and capabilities that students are expected to develop during their programme of study, increasingly including AI literacy. Read more on the Quality page.
H
- Higher-Order Thinking Skills
- Cognitive abilities including analysis, evaluation, synthesis, and creation. Assessment redesign prioritises these skills as they are less susceptible to AI automation. Read more on the Framework page.
I
- Institutional Supports
- Resources and structures at faculty, institutional, and sectoral levels that enable sustainable assessment redesign, including professional development, cross-disciplinary dialogue, and unified guidance. Read more on the Quality page.
L
- Learning Outcomes
- Statements describing what learners are expected to know, understand, and be able to do. The central guiding principle of assessment redesign is alignment with these outcomes. Read more on the Quality page.
M
- Multi-Stage Assessment
- Assessment design where students submit plans, drafts, checkpoints, or progress updates before a final submission, making the evolution of their thinking visible. Read more on the Strategies page.
O
- Oral Components
- Assessment elements such as presentations, vivas, demonstrations, or Q&A discussions that enable students to explain, justify, and reflect on their work. These are lower risk for AI misuse. Read more on the Risk & Detection page.
P
- Postplagiarism
- A concept that encapsulates the new challenges to academic integrity as AI tools become integrated into academic work, requiring re-evaluation of traditional definitions of plagiarism. Read more on the Resources page.
- Process-Based Assessment
- Assessment that emphasises how learners arrive at conclusions, not just final products, making learning visible through reflective journals, drafts, staged tasks, and iterative feedback. Read more on the AI Integration page.
- Programme-Level Assessment Mapping
- The review of assessment patterns across all stages of a programme to ensure a balanced mix of assessment types and consistent expectations around AI use. Read more on the Quality page.
Q
- Quality Assurance
- Institutional processes ensuring assessment redesign maintains fairness, coherence, and academic standards, embedded within programme review, annual monitoring, and curriculum approval workflows. Read more on the Quality page.
R
- Reflective Elements
- Assessment components requiring students to articulate their reasoning, learning processes, and how they used or chose not to use AI tools. Read more on the Strategies page.
- Risk Assessment
- The evaluation of how vulnerable different assessment types are to inappropriate GenAI use, categorised as high, medium, or low risk. Read more on the Risk & Detection page.
S
- Structural Redesign
- Changing the mechanics of assessment to build validity into the task itself, as opposed to discursive control which relies on rules and student compliance. Read more on the AI Integration page.
- Summative Assessment
- Assessment designed to evaluate and grade student learning at the conclusion of a period of study. Redesign aims to balance summative with formative approaches. Read more on the Strategies page.
T
- Track 1 (AI-Restricted / Secured)
- Assessment where AI use is typically not permitted. The focus is on 'assessment of learning' through supervised conditions. Unauthorised AI use is considered a breach of academic integrity. Read more on the AI Integration page.
- Track 2 (AI-Integrated / Open)
- Assessment where responsible use of AI is encouraged. These are less supervised, promoting engagement with AI and preparing students for an AI-integrated society. Acceptable usage is detailed in Assessment Briefs. Read more on the AI Integration page.
- Transparency
- Clearly communicating expectations around the purpose of assessment, criteria for success, and the extent to which AI use is permitted or restricted. One of the five core principles. Read more on the Principles page.
U
- Universal Design for Learning (UDL)
- A framework for designing inclusive educational experiences that offer diverse ways for students to engage with content and demonstrate their learning. Read more on the Principles page.
V
- Validity
- The principle that assessment tasks genuinely measure the knowledge, skills, and competencies they are intended to assess, even in contexts where AI tools are widely available. The first core principle. Read more on the Principles page.