How does AI assessment redesign help beginners?
Beginners primarily benefit from AI assessment redesign through structured AI literacy education and clear guidelines. In a 2025 study of a Swedish physiotherapy program, faculty implemented targeted AI literacy training and redesigned assessments to emphasize clinical reasoning and critical thinking, reducing opportunities for misuse [1]. Two months after implementation, 42% of students felt adequately informed about AI, and faculty AI literacy and confidence improved through structured group work [1]. This shows that beginners need explicit instruction on when and how to use AI, not just redesigned tasks.
Another 2025 study of 61 faculty members found that key motivations for redesigning assessments included maintaining academic integrity and preparing learners for future careers, but significant challenges remained, such as the need for professional development and addressing equity and accessibility concerns [5]. This suggests that for beginners, redesign alone is insufficient without accompanying training and support.
How does AI assessment redesign benefit experts?
Experts benefit from AI assessment redesign by advancing their integration of AI into professional practice. A 2025 study developed and validated an AI-TPACK assessment tool using 60 authentic teaching artifacts from teacher educators [2]. The tool identified four competency patterns: technological innovator, pedagogical integrator, content developer, and beginner. The strong correlation (r = 0.78) between AI pedagogical knowledge and integration underscores that experts can use redesign to deepen synergy between AI knowledge and classroom practice [2].
A 2024 study of master's students found that many existing assignments could be partly solved with generative AI, and students recommended continuous assessment and oral examinations as complements [4]. This implies that experts—both faculty and advanced students—can leverage redesign to create more resilient assessments that test deeper understanding, not just recall.
What works for both beginners and experts?
Across the studies, two strategies consistently benefit both groups: clear guidelines on acceptable AI use and continuous assessment formats. The 2025 physiotherapy study integrated standardized guidelines across all courses, reducing misuse [1]. The 2024 master's student study recommended clear instructions about when AI is allowed and redesigning course structure for continuous assessment, where the whole study path is assessed, not just isolated submissions [4]. This approach strengthens fairness and sustainability for all learners.
A 2026 paper introduced the AI Risk Tensor (ART), a three-dimensional framework adding 'Opacity' as a risk dimension alongside Impact and Probability [3]. While focused on organizational risk, this concept applies to assessment: both beginners and experts need to understand the opacity of AI outputs (e.g., why an AI gave a certain answer) to use AI responsibly. The 2025 faculty study's 'Against, Avoid, Adopt, and Explore' framework [5] similarly provides a spectrum of responses that can be tailored to different expertise levels.
About These Sources
This answer is built on 5 peer-reviewed studies — published from 2024 to 2026, 5 from 2024 or later, 2 in Q1 journals — selected as the most relevant from 5 studies that passed quality screening, drawn from 56 papers retrieved from a database of over 500 million.
Sources used in this answer
Structured Development of Learning and Assessment Tasks to Prevent Generative AI Misuse and Enhance AI Literacy in the Faculty in Physiotherapy Education
In a Swedish physiotherapy program, faculty redesigned assessments to emphasize clinical reasoning and critical thinking; 42% of students felt adequately informed about AI after implementation, and faculty AI literacy improved through structured group work [1].
Developing and Validating an AI-TPACK Assessment Framework: Enhancing Teacher Educators’ Professional Practice Through Authentic Artifacts
A validation study of 60 authentic teaching artifacts from teacher educators identified four AI competency patterns (technological innovator, pedagogical integrator, content developer, beginner) and found a strong correlation (r = 0.78) between AI pedagogical knowledge and integration [2].
Redesigning Risk Assessments in the Age of AI: Emergence of the Opacity Risk Dimension.
A 2026 paper introduced the AI Risk Tensor (ART), a three-dimensional framework adding 'Opacity' as a risk dimension alongside Impact and Probability, to address unpredictability in AI systems [3].
Generative AI and its Impact on Activities and Assessment in Higher Education: Some Recommendations from Master's Students
A study of 16 master's students found that many assignments could be partly solved with generative AI; students recommended clear guidelines on AI use and continuous assessment formats, including oral examinations [4].
Redesigning Assessments for AI-Enhanced Learning: A Framework for Educators in the Generative AI Era
A qualitative study of 61 faculty members identified motivations for redesigning assessments (academic integrity, career preparation) and challenges (professional development needs, equity concerns), leading to the 'Against, Avoid, Adopt, and Explore' framework [5].