YOUR GO-TO INTELLIGENCE RESOURCE FOR THE SCHOLARLY PUBLISHING INDUSTRY

Initial insights from AI in marking and feedback pilots conducted across 38 UK colleges and universities reveal both opportunities and challenges in applying AI tools to assessment. The pilots, which involved Graide, Keath, and TeacherMatic, created a collaborative environment for educators and students to explore how AI might reduce workload while maintaining academic standards.

The pilot model allowed participants of varying skill levels to share experiences and identify areas where caution is necessary. Supported by sector experts and product developers, the initiative provided a safe space to examine how AI could be integrated into institutional practices.

Key findings emphasized the importance of keeping human oversight in the loop, with parallel marking identified as an effective method. Open dialogue and clear communication channels were found to help secure student buy‑in. Participants also noted that AI should not be applied universally; instead, evidence‑based rationales are needed to justify its use where efficiency gains can be demonstrated without compromising integrity. Formative assessment emerged as the most effective and low‑risk application of AI in marking and feedback.

Additional outcomes included opportunities for collaboration between participants and developers, with user feedback informing tool improvements. Shared challenges included communicating AI use to staff and students, which led to the creation of resources such as a student communication pack. Participants also observed that clearly defined marking criteria improved the accuracy and consistency of AI outputs.

Tom Moule, senior AI specialist at Jisc, explained that the pilot focused on fostering informed discussion and a collaborative approach to determining where AI could be beneficial and where caution was required. He noted that conversations often shifted from the tools themselves to broader questions about assessment practices.

Hilary Reid, digital learning team leader at Ayrshire College, highlighted that peer support from other institutions was one of the main benefits of participation. She added that the pilot provided direct feedback channels to developers, resulting in improvements to TeacherMatic that benefited users. Reid also observed that structured sector‑wide participation gave staff confidence to explore AI tools in their own practice, leading to increased adoption at Ayrshire College.

As the pilots continue, Jisc’s AI team plans to use the findings to support UK further and higher education in developing scalable approaches that balance the benefits of AI with the need to safeguard trust, quality, and professional judgment.