Development and Evaluation of an Educational Conversational Agent for Promoting Deepfake Literacy
Status
Dissemination phase
Acronym
Deepfake Literacy
Project contact PHSG
Lay Summary
The rapid advancements in artificial intelligence (AI) have led to an increase in the production and dissemination of deepfakes—AI-generated multimedia content that manipulates video, audio, or images in a realistic manner. Deepfakes pose a significant threat to digital literacy, democratic processes, and societal trust, particularly as they are increasingly used to spread misinformation. Despite growing awareness of conventional disinformation, there is a substantial gap in educational tools designed to promote deepfake literacy. This project aims to address this gap by developing and evaluating an educational conversational agent (chatbot) to help students and educators recognize and understand deepfakes.
The agent will serve two main functions: first, to deliver instructional content on identifying deepfakes, and second, to offer interactive tutoring that reinforces learning through simulated scenarios. For instance, learners will evaluate different deepfakes and assess their truth factors, receiving personalized tutoring and adaptive feedback based on their responses. Conversational agents have demonstrated effectiveness in educational contexts by promoting engagement, providing immediate feedback, and personalizing learning experiences at scale. Leveraging this technology, we aim to enhance learners’ ability to critically assess multimedia content, fostering deeper media literacy and critical thinking skills.
The project follows the Educational Design Research (EDR) methodology, involving iterative cycles of design, testing, and evaluation. The initial phase includes a comprehensive literature review and participatory workshops with educators to identify key challenges in teaching deepfake literacy. These insights will inform the design and development of a conversational agent prototype, which will be tested in real-world classroom settings. Data collected through mixed-methods evaluation—including quantitative measures such as usability and cognitive load, and qualitative feedback from teacher interviews—will guide further development.
The collaboration between Bern University of Applied Sciences (BFH) and the University of Teacher Education St.Gallen (PHSG) ensures a multidisciplinary approach, combining expertise in educational technology and pedagogical practices. This project aligns with the goals of BeLEARN by addressing a pressing issue in the digital and AI-driven world. The long-term vision is to expand the agent’s application to broader contexts of digital literacy, securing additional funding from the Swiss National Science Foundation (SNSF) or the Hasler foundation to further enhance its capabilities and impact.
The agent will serve two main functions: first, to deliver instructional content on identifying deepfakes, and second, to offer interactive tutoring that reinforces learning through simulated scenarios. For instance, learners will evaluate different deepfakes and assess their truth factors, receiving personalized tutoring and adaptive feedback based on their responses. Conversational agents have demonstrated effectiveness in educational contexts by promoting engagement, providing immediate feedback, and personalizing learning experiences at scale. Leveraging this technology, we aim to enhance learners’ ability to critically assess multimedia content, fostering deeper media literacy and critical thinking skills.
The project follows the Educational Design Research (EDR) methodology, involving iterative cycles of design, testing, and evaluation. The initial phase includes a comprehensive literature review and participatory workshops with educators to identify key challenges in teaching deepfake literacy. These insights will inform the design and development of a conversational agent prototype, which will be tested in real-world classroom settings. Data collected through mixed-methods evaluation—including quantitative measures such as usability and cognitive load, and qualitative feedback from teacher interviews—will guide further development.
The collaboration between Bern University of Applied Sciences (BFH) and the University of Teacher Education St.Gallen (PHSG) ensures a multidisciplinary approach, combining expertise in educational technology and pedagogical practices. This project aligns with the goals of BeLEARN by addressing a pressing issue in the digital and AI-driven world. The long-term vision is to expand the agent’s application to broader contexts of digital literacy, securing additional funding from the Swiss National Science Foundation (SNSF) or the Hasler foundation to further enhance its capabilities and impact.
Relation to the professional field
Development project
Topic PHSG
Digitale und Informatische Bildung
Fields of Science and Technology (OECD)
Educational sciences
Psychology (including human - machine relations)
Project contact PHSG
Consortium Coordinator
Partner Organisations
Berner Fachhochschule
Consortium Coordinator(s)
Principal Investigator
Investigators
Co-Investigator(s) external
Wambsganss, Thiemo
Müller, Livia
| Funder | Start | End | Award URL Funder | Grant Number Funder | Funder Program |
January 2025 | June 2025 | BeLEARN Booster | |||
Note on the external funding