Creating a Culture of Trust in Teaching Academic Writing for Foundation Year Students through Personalised AI Support

Generative AI poses distinct challenges for Foundation Year students, who are simultaneously developing academic literacies, self-efficacy, and understanding of scholarly integrity. While AI tools offer personalised feedback and linguistic scaffolding, they also generate uncertainty about authorship boundaries, inconsistent guidance across courses, and erosion of trust among students, instructors, and institutions. This paper argues that cultivating trust is the essential precondition for meaningful AI integration in academic writing pedagogy. Drawing on academic literacies research, trust theory in education, and critical AI studies, this paper proposes a three-part framework: (1) trust-building pedagogical practices emphasising transparency, dialogue, and psychological safety; (2) coherent institutional communication that resolves policy ambiguity around ethical AI use; and (3) personalised AI writing support designed to scaffold rather than substitute learning. This framework is illustrated through a three-day workshop at Oxford University with multilingual students (aged 18+) preparing for English-speaking universities, employing a proprietary AI tool. Analysis of student reflections reveals how transparent discussion of AI's pedagogical constraints fostered trust-based engagement. Unlike compliance-focused approaches positioning AI as a threat, this framework reconceptualises integration as fundamentally a problem of pedagogical relationships and institutional culture. By positioning trust as a prerequisite rather than an outcome, it offers an urgently needed perspective on how higher education can respond to generative AI not as a technological challenge but as an opportunity to strengthen the relational and ethical dimensions of writing pedagogy.