AI in Education Bi-weekly Digest - Issue #25

Welcome to our latest #25 newsletter. In this issue, we bring you noteworthy events from educators worldwide, including highlights from the LEADING-AI forum organized by The Hong Kong Polytechnic University. Join us as we explore innovative strategies from industry experts—from Deloitte’s insights on leveraging generative AI in higher education to discussions on the challenges of maintaining academic integrity in an AI-driven world. We also celebrate groundbreaking publications and cutting-edge tools that are transforming teaching, learning, and research. Dive in to discover how educators are shaping the future of education in the age of AI.

Noteworthy Events from Educators Worldwide:

LEADING-AI - The "Learning Design, Innovation, and Generative AI: Virtual Special Forum" has successfully concluded. Organized by The Hong Kong Polytechnic University, the event featured a dynamic presentation by Professor Hassan Khosravi and an engaging panel discussion with leading experts. Together, they explored how AI is transforming educational practices, offering fresh strategies to enhance teaching and learning. Recording and slides are available here.

News, Articles & Blogs:

[Sharing] How Higher Education Can Realize the Potential of Generative AI - Deloitte examines how higher education leaders can harness the transformative power of generative AI to drive innovation across teaching, research, and operations. Drawing on insights from industry experts at the Deloitte Center for Government Insights, it highlights the need for a culture of change—where students, faculty, and staff overcome apprehensions to embrace AI as a collaborative tool. In a succinct 20-minute read, the piece outlines strategies for safely integrating AI to enhance efficiency, effectiveness, and overall academic excellence.

[Comment] AI-detection Software Isn't the Solution to Classroom Cheating - Assessment Has to Shift - Traditional take-home assessments no longer reliably reflect what students truly know since they are vulnerable to various cheating methods. Instead of investing in imperfect detection technology, the focus must shift toward authentic, ongoing formative assessments that gauge real learning and emphasize academic integrity as a shared responsibility among students, teachers, and institutions.

[Report] HEPI/Kortext AI Survey Shows Explosive Increase in the Use of Generative AI Tools by Students - The latest AI survey reveals a staggering surge in generative AI use among undergraduates—with usage for assessments soaring from 53% to 88% in just one year. Authored by Josh Freeman, the report highlights not only the rapid adoption of tools like ChatGPT for explaining concepts and generating research ideas but also draws attention to challenges such as academic integrity and a growing digital divide. With calls for bold policy updates and collaborative best practices, the study underscores that AI is now a core element of the higher education landscape. The report can be found here.

[Publication] Onto-Epistemological Understandings of Generative Artificial Intelligence in Education - Congratulations to our esteemed former keynote speaker on the publication of "Onto-Epistemological Understandings of Generative Artificial Intelligence in Education" in the International Journal of Changes in Education. This thought-provoking article by Edwin Creely and Kitty Janssen brilliantly explores the evolving relationship between humans and GenAI, offering fresh, groundbreaking perspectives that promise to reshape educational paradigms. We are proud to celebrate this significant achievement and look forward to the continued impact of your visionary work in advancing teaching and learning in the AI era.

[Publication] Developing and validating measures for AI literacy tests: From self-reported to objective measures - Another congratulation to our conference speaker, Prof Thomas Chiu, on the publication of "Developing and validating measures for AI literacy tests: From self-reported to objective measures." This significant work—spanning interdisciplinary collaboration and innovative test development—marks a milestone in accurately assessing AI literacy among school students. By shifting from self-reported data to objective measures and rigorously validating these through a robust Rasch model, his research lays a strong foundation for future AI education practices.

[News] Alibaba Makes AI Video Generation Model Free to Use Globally - Alibaba has announced that its latest video generation AI models from the Wan2.1 series are now free and open source for global use. These models, which generate images and videos from text and image inputs, are available via Alibaba Cloud’s Model Scope and the Hugging Face repository, making them accessible to researchers, academics, and commercial institutions worldwide. This move follows the recent trend by Chinese firms to bolster innovation through open-source AI, underscoring Alibaba’s commitment to advancing technology while fostering a collaborative community.

[Comment] UK AI Copyright Policy Is Wrong And Based on Fading Ideas - British creative industries are voicing concerns as the government debates amending copyright law to favor AI developers. Dr Dominic Lees from the Synthetic Media Research Network at the University of Reading criticizes the proposal, arguing it relies on outdated models of "scraping" copyrighted content. Instead, he advocates for a centralized database of approved creative works, enabling AI developers to legally license content. Lees contends this forward-thinking approach would protect intellectual property, offer fair remuneration, and boost the UK's AI competitiveness, while aligning with international copyright treaties.

[Sharing] Think You Can Cheat with AI? A UF Professor Creates Watermarks to Detect AI-Generated Writing - A University of Florida engineering professor, Yuheng Bu, is developing an invisible watermarking technique to detect AI-generated text—even when modified. Using UF’s HiPerGator supercomputer, his team embeds imperceptible signals into a subset of text produced by large language models, ensuring high text quality and robust detection even after paraphrasing. The watermark relies on a private key held by the watermarking entity for verification. Bu envisions this method as a valuable tool for confirming authenticity in academic and professional settings, ultimately reducing AI-assisted cheating while fostering trust in digital content across various platforms.

[Sharing] 5 New Trends in Generative AI That Web3 Needs to Be Ready For - This author, Jesus Rodriguez, unveils five groundbreaking trends in generative AI—from enhanced reasoning to compact, efficient distilled models—and explores how these innovations open new doors for Web3 integration. Authored by Jesus Rodriguez and edited by Nelson Wang, it examines how decentralized technologies can bring transparency, efficiency, and collaboration to AI development. Dive in to discover how Web3 can be future-ready in the evolving AI landscape.

Tools/Resources:

Disclaimer: AI in Education has no affiliation with any highlighted free or commercial products in this section.

[Tool] Simulating Scientist: A New Tool for AI-Powered Scientific Discovery - Monash University researchers have unveiled LLM4SD—a groundbreaking generative AI tool that simulates scientific discovery. By reading decades of scientific literature and analyzing lab data, LLM4SD can retrieve information, form hypotheses, and predict molecular behaviors across diverse fields such as physiology, chemistry, biophysics, and quantum mechanics. In contrast to many "black box" models, this interactive system explains its reasoning process using simple, transparent rules, boosting prediction accuracy by up to 48% over traditional tools. Freely available and open source, LLM4SD can be accessed here.

---

Disclaimer: The views and opinions expressed in the linked posts are those of the speakers and or their entities and do not necessarily reflect the views or positions of the project AI in Education, Centre for Learning Enhancement And Research, and The Chinese University of Hong Kong.