Gaming Ethics Across Borders: Exploring Responsible AI Through Play

Summary

The project built on the previous RAi UK–funded Exploring AI: From Picture Books to Epic Games initiative by shifting the focus from educational resource development towards the use of megagames as empirical research tools for Responsible AI. 

The project aimed to: 

• Explore how immersive, game-based experiences shape understanding of Responsible AI concepts such as trust, fairness, accountability, human oversight, and bias
• Compare UK and Australian perspectives on AI-enabled decision-making and information processing
• Develop a methodological framework and future research agenda for using games as tools for Responsible AI skills development, public engagement, and behavioural research

Findings will feed into a practical framework showing how games can be used to build Responsible AI skills and support future public engagement and education in the UK, Australia and beyond.

Collaborating with Norwich University of the Arts, University of Tasmania.

Outputs 

The project delivered the following outputs: 

• A research-adapted version of the RAi-funded AI megagame suitable for empirical study
• Two comparable UK–Australia datasets combining qualitative and quantitative data
• Shared workshop protocols and data collection instruments
• Foundational methodological principles and considerations for future frameworks exploring how games can support Responsible AI engagement and research
• A strengthened international collaboration between UK and Australian researchers spanning cybersecurity, cognitive psychology, media psychology, and game design
• A future research agenda targeting larger-scale interdisciplinary and international collaboration. 

Activities and Achievements

Refinement of the RAi UK Megagame

The team adapted the existing RAi-funded megagame into a research-ready format suitable for empirical data collection. Scenarios were refined to foreground ethically challenging AI-related decision-making situations involving misinformation, fairness trade-offs, trust calibration, and AI-supported recommendations. 

UK and Australian Workshops

Linked workshops were delivered in the UK and Australia using the same megagame structure and scenarios. Participants included students, researchers, and practitioners with interests in AI, digital technology, and behavioural science. 

The game-based format enabled participants to engage with Responsible AI concepts in an interactive, social, and highly applied environment rather than through traditional lecture or survey-based methods. The workshops also created opportunities for structured reflection on the ethical and societal implications of AI, supporting Responsible Research and Innovation principles through participatory and discussion-led engagement. 

Collaborative Analysis and Early Framework Development

Researchers from all partner institutions conducted comparative analysis across both sites, identifying common themes and notable differences in how participants reasoned about AI-enabled systems. 

The collaboration also initiated development of a conceptual framework for Exploring AI Through Play. Rather than producing a finalised model, the project identified key methodological considerations, Responsible AI challenges, and opportunities for future interdisciplinary research using immersive game-based methods. Discussions focused particularly on how collaborative play environments can surface attitudes towards fairness, trust, responsibility, and human oversight in AI-supported decision-making. 

Impact

The project demonstrated that megagames and other large-scale collaborative play environments have significant potential as Responsible AI engagement and research tools. Participants consistently reported that the immersive and social nature of the game enabled them to think more critically about trust, fairness, accountability, and human oversight in AI-enabled systems.