As artificial intelligence becomes increasingly integrated into educational assessment, organisations face a critical challenge: how to harness the benefits of AI while preserving the value of human expertise, judgement, and accountability. At Cambridge University Press & Assessment, we have developed a practical, principles-led approach to AI adoption that places people at the centre of every stage of the assessment lifecycle.

This presentation explores how our Ethical AI Principles have moved beyond policy and into practice. We begin by outlining the principles that guide our use of AI and the governance structures that ensure they remain embedded in decision-making. We then present a case study of our AI-assisted assessment content creation tool, demonstrating how human expertise informed its design, development, and implementation. From defining requirements and training processes to reviewing outputs and safeguarding quality, subject matter experts have remained central to the solution, ensuring that AI acts as an enabler rather than a replacement for professional judgement.

Looking beyond the technology itself, we discuss how we are preparing our assessment teams for a future in which AI becomes part of business-as-usual operations. This includes redesigning content creation pathways, reconfiguring team roles and responsibilities, and establishing clear decision points where human oversight and expertise are essential. We also examine our approach to capability building, including how we train colleagues to use AI ethically and effectively, how we evaluate the benefits and risks of AI-assisted processes, and how we maintain trust, quality, and accountability throughout the content development lifecycle.

The session will provide a practical example of responsible AI implementation within a high-stakes assessment environment and offer insights for organisations seeking to embed AI sustainably while ensuring that human expertise remains at the heart of innovation.