A provocation: the manifesto for feedback

20 February 2026

In this post, CRADLE PhD student Siham AbuKhalaf reflects on the timely and salient messages of our first seminar for 2026: Manifesto for feedback in the age of AI. The topics are closely aligned with Siham’s own research areas whereby she is investigating the relationship between AI, authentic assessment, and student motivation.

Manifesto for feedback in the age of artificial intelligence

CRADLE’s seminar on the Manifesto for feedback in the age of artificial intelligence centred on a pressing question: as generative AI becomes embedded in higher education, what must we hold on to in feedback practice?

Professor Naomi Winstone introduced the manifesto as a principled response to uncertainty and outlined its core principles:

feedback is a process, not comments

it is a relational practice

it can be messy, uncomfortable, challenging and sometimes joyous

it must remain ethical, grounded in respect, transparency and equity

and it should promote learning over time, rather than prioritising speed, efficiency or scale

Naomi also emphasised respect for scholarship, professional craft, human connection and equity as essential anchors when making decisions about AI integration. The manifesto, she argued, is not anti-AI; instead, technology should be judged against what feedback scholarship has already established as valuable.

The panel then tested these principles through a medical education scenario in which an AI-enabled virtual patient provides students with instant diagnoses and polished summaries before tutor debriefs. Associate Professor Christy Noble raised concerns about what tutors lose when students’ authentic reasoning pathways are replaced by reflections shaped by AI output. If feedback is a process, what happens when that process is pre-empted?

Associate Professor Kelli Nicola-Richmond reflected on principle three—the productive role of discomfort and messiness in learning. She noted that struggle often drives transformation, yet also acknowledged that for some learners, carefully used GenAI might reduce overwhelm and enable entry into meaningful learning. The question becomes not whether to remove struggle, but how much struggle promotes learning and for which students.

Dr Anna Jones approached the scenario through the ethical principle. She highlighted potential benefits such as reduced bias, safe rehearsal spaces, and diminished power imbalances, while questioning whether AI-mediated encounters can cultivate care, nuanced clinical judgement and the complexity of human interaction.

Associate Professor Karen Gravett drew the discussion together by identifying the “thorny questions” that the manifesto helps surface, namely: how do we ensure learning, not compliance or efficiency, drives design? What skills might be eroded? What affordances might emerge?

The seminar ultimately positioned the manifesto as a provocation.

In an AI-rich landscape, it invites us to continually ask: does this design strengthen feedback as relational, ethical and developmental practice or does it quietly reduce it to information?

Access the manifesto now

About Siham

Siham AbuKhalaf is a PhD student with CRADLE. Her research is focusing on the intersection between authentic assessment, GenAI, and student motivation in higher education. Siham recently earned her master’s degree in educational leadership and policy and is a TEDx speaker, an active researcher with several publications, and a contributor to multiple academic conferences.