Keynote: Transforming Language Assessment with AI-Powered Innovations
For most of the past two decades, large language testing companies have maintained a significant technological advantage over university departments and smaller organizations. With access to interdisciplinary teams spanning language assessment, computer science, and artificial intelligence, large-scale test publishers have led advancements in areas such as automated item generation (AIG), automated essay scoring (AES), and automated speech assessment. In contrast, university researchers and developers of local language assessments have often lacked comparable resources, and have, arguably, lagged behind in these critical areas.
However, recently we have witnessed a seismic shift in the technological landscape. The democratization of AI and computational resources — through publicly available massive corpora, affordable and scalable cloud computing (e.g. AWS), off-the-shelf foundation models (e.g. AWS Bedrock), deep learning libraries (e.g. TensorFlow), and open-source large language models (e.g. LLaMA, DeepSeek) — has made sophisticated tools more accessible than ever before. These developments present an unprecedented opportunity for university-based researchers, long proficient in advanced statistical and psychometric analyses, to reclaim leadership in language assessment technology and innovation.
This presentation examines the skill sets, tools, and strategies that university researchers must adopt to harness these advancements. As institutions increasingly develop local language assessments tailored to their specific contexts, embracing these technologies can not only modernize test development, administration, and validation but also reposition universities to the forefront of transformative change in language assessment.
Alistair Van Moere
Alistair is President at MetaMetrics Inc, and Research Professor at The University of North Carolina at Chapel Hill. He drives innovation in educational AI and assessments, guarantees the validity of test scores for 35 million students every year, and helps over 100 organizations make sense of educational metrics. Previously, he was President of the division in Pearson that provided AI scoring services for dozens of assessments and speaking and writing programs. He also oversaw development and delivery of standardized assessments for tens of millions of learners. Alistair has worked as a teacher, examiner, director of studies, university lecturer, and EdTech executive, in the US, UK, Japan, and Thailand. He has an MA in English Language Teaching and MBA from Warwick Business School. His PhD from Lancaster University won the Jacqueline Ross TOEFL award for best dissertation in language testing. He is frequently an invited speaker at conferences and has authored over 20 research publications with a focus on educational technology and language assessments.
