🧠 AI Ed Wiki

📄 Research Article

Feedback futures: beyond the limits of human and GenAI capacities

● high · 2026-08-03generative-ai, feedback-loop, higher-ed, ai-literacy, educational-theory, equity, equity-in-ai-education, human-in-the-loop-ai, rag, ai-education

Summary

This editorial synthesises the seven papers of the AEHE 51(5) special issue on feedback in the age of generative AI. Its central claim: the question is not whether GenAI feedback is useful, but how human and GenAI feedback can be combined to sustainably support learning rather than merely improve immediate performance. Teacher and student feedback literacy are necessary but not sufficient — what is also required is deliberate pedagogic design, institutional conditions that prioritise learning over efficiency, and the purposeful development of human judgement that cannot be delegated to the tool.

The issue's papers include four also ingested into this wiki: Learner Centered Feedback AI, Chatgpt Feedback Engagement GenAI, GenAI Teacher Feedback Comparison, and Care Full Feedback GenAI.

Five tensions in GenAI feedback

The editors distill five recurring tensions from the special issue:

1. Usefulness, trust, and uptake. GenAI is valued for speed, clarity, and accessibility, while teacher feedback is trusted for contextual understanding, disciplinary expertise, accountability, and human connection. Uptake depends on experienced care, recognition, and presence — relational conditions GenAI struggles to reproduce.

2. Immediate task achievement vs longer-term learning. GenAI excels at helping students complete the task at hand but may orient them toward performance/avoidance goals rather than mastery — echoed in Chatgpt Feedback Engagement GenAI's finding of weak metacognitive engagement and one-off interactions.

3. Agency vs dependency. Agency can be extended via iterative prompting and comparison, but may become "thinner" when students stay active at the level of interaction while ceding evaluative work to the system.

4. Access, avoidance, and advantage. GenAI is unlikely to benefit all students equally; non-users cite trustworthiness, preference for human feedback, and academic integrity. Access alone does not guarantee educative uptake.

5. Teacher judgement and labour redistribution. GenAI redistributes rather than removes teacher labour — teachers still assess accuracy, tone, relationality, and pedagogical value, and poorly designed tools can increase workload (see Learner Centered Feedback AI).

Conditions of partnership

  • GenAI and human feedback are complementary but only if the complementarity is specified — via sequencing, comparison, editing, and governance across multiple sources and moments ("an ecology of configured encounters").
  • The subtler risk is displacement, not replacement: the speed of GenAI feedback may gradually displace the slower, careful work of educational judgement without anyone consciously deciding it should.
  • Human agency must be proactively maintained rather than assumed — humans remain responsible for judging, adapting, and rejecting GenAI responses.
  • The editors warn against both "AI slop" and "human slop" (low-quality, standardised feedback from comment banks and templates), arguing GenAI may entrench the latter if ungoverned (see Care Full Feedback GenAI).

    • Feedback literacy for GenAI contexts

  • In a GenAI-enhanced essay-writing study in the issue, feedback literacy predicted performance while frequency of GenAI use, trust, and prior knowledge did not — the learner's capacity to seek, evaluate, and act on feedback shapes GenAI's educational value.
  • Existing feedback literacy frameworks (Carless & Boud 2018; Molloy et al. 2020) must be extended with GenAI-specific capacities: evaluative judgement (GenAI output can look authoritative while being hallucinatory), metacognitive skill (monitoring when GenAI supports vs narrows vs substitutes thinking), and ethical decision-making (when and how GenAI use supports rather than substitutes for students' own intellectual work).
  • Teacher feedback literacy (design, relational, pragmatic dimensions; Carless & Winstone 2023) is less well theorised: design requires workable GenAI+human feedback workflows; relational oversight cuts both ways (teacher feedback is often perceived as more negative/risky than GenAI — see GenAI Teacher Feedback Comparison); pragmatically, what must remain human-led is not only connection but accountable judgement.

    • Governance and assessment design

  • Policy cannot be purely top-down: acceptable GenAI support varies across courses and disciplines, and ambiguous institutional direction pushes responsibility down to individual teachers, individualising the challenge.
  • Assessment reform should make learning processes visible rather than treating final products as direct evidence; institutions must be deliberate about when GenAI use is itself part of what is being assessed.
  • Research should move beyond self-report toward in-situ methods (think-aloud, trace/log data, stimulated recall) and toward how learners orchestrate feedback across sources (human, GenAI, artefactual) that differ in credibility and value. Equity research should track not just tool access but educationally productive use.

    • Connected Concepts

  • AI Literacy
  • Equity In AI Education
  • Higher Ed
  • Human In The Loop AI
  • AI Education
  • Generative AI
  • RAG

    • Connected Articles

  • Agency Gap AI Writing — The agency gap in AI-supported writing: how reactive and proactive agent designs shape multimodal reasoning
  • Care Full Feedback GenAI — The care-full craft of feedback in an age of generative AI
  • Chatgpt Feedback Engagement GenAI — Students' engagement with ChatGPT feedback: implications for student feedback literacy in the context of generative a...
  • GenAI Teacher Feedback Comparison — Comparing Generative AI and teacher feedback: student perceptions of usefulness and trustworthiness
  • Learner Centered Feedback AI — Enhancing learner-centered feedback with AI: teachers' practices and perceptions
  • A4l Analytics Pipeline — Generalizing a Highly Configurable Analytics Pipeline to Replicate and Support Educational Research Across Multiple D...
  • Aaai2026 Prompting Literacy K12 — Learning to Use AI for Learning: Teaching Responsible Use of AI Chatbot to K-12 Students Through an AI Literacy Module
  • Academiclaw Student Agent Benchmark — AcademiClaw: When Students Set Challenges for AI Agents
  • Access Not Enough AI Tutoring 2026 — Access is Not Enough: Human Support Improves Engagement with AI Tutoring
  • Adapt Adaptive Lesson Plan Transformer — AdaPT: Adaptive Lesson Plan Transformer for Cross-Regional and Differentiated Instruction
  • Adaptive Pretesting Retention — Do Gains from Generative AI-Enabled Adaptive Pretesting Persist? Evidence from a Retention Study
  • Affective Text Wearable Student Health — A Formative Study of Brief Affective Text as a Complement to Wearable Sensing for Longitudinal Student Health Monitoring
  • Agent Voice Accents K12 Group Learning — Exploring How Agent Voice Accents Shape Human-AI Collaboration in K-12 Group Learning
  • Agentic AI Education Scoping Review — Agentic AI in Education: A Scoping Review of Research Landscape, Capabilities, and the Frontier Agent Paradigm
  • Agentic AI Pedagogical Best Practice 2026 — Agentic AI and Pedagogical Best Practice: The Tension Between Automation and Learning
  • Agentic Education Coding — Agentic Education with AI Coding Assistants
  • Agentic Literacy Debt — Agentic Literacy Debt: A Structural Problem the AI Literacy Field Has Not Yet Named
  • Agentic Workflows Education — Agentic Workflows in Education
  • Agents That Teach Incidental Learning — Agents That Teach: Designing Incidental Learning Back into AI-Assisted Software Development
  • Agreement Not Quality LLM Coding Verification — Agreement Is Not Quality: Blind Expert Verification of Human and LLM Qualitative Coding When Human Consensus Is Not G...
  • AI Adoption Training Public Sector — The Main Barrier to AI Adoption in the Public Sector is Lack of Training
  • AI Adult Learning Guidelines Dis2026 — Guidelines for Designing AI Technologies to Support Adult Learning
  • AI Agents Constructive Conflict Design Education 2026 — Enacting Constructive Conflicts with AI Agents to Enhance Reconsideration among Novice Interaction Designers
  • AI Assessment Human Tutors — AI-Driven Assessment of Human Tutors: Linking Training Performance to Real-Life Practice
  • AI Assessment Scale Reform — A bit of chaos and madness": The AI Assessment Scale and the work of assessment reform

    • Citation

    Zhan, Y., Wood, J., Carless, D., & Yan, Z. (2026). Feedback futures: beyond the limits of human and GenAI capacities. Assessment & Evaluation in Higher Education, 51(5), 811–820