Search papers, labs, and topics across Lattice.
This paper details the implementation of "Capture the Narrative," a four-week competition where student teams from 18 Australian universities developed LLM-powered bots to manipulate a simulated election, addressing the inadequacies of traditional digital literacy education. Despite the high engagement, with over 7 million posts generated, the results revealed that students did not gain confidence in identifying misinformation, contradicting expectations based on inoculation theory. The findings highlight the challenges of fostering effective cyberliteracy in a landscape increasingly dominated by AI-generated content and suggest the need for more dynamic educational approaches.
Students engaged in a bot-driven election simulation failed to improve their ability to spot misinformation, challenging the effectiveness of traditional digital literacy strategies.
Misinformation is deeply embedded in online discourse, with nearly one in five posts during global events generated by bots that amplify false content. In recent years, the use of Generative AI has further lowered the barrier to producing convincing misinformation, yet most digital literacy education still relies on static checklists and single-player inoculation games built for an earlier media landscape. This paper describes how we addressed this educational gap through Capture the Narrative, a four-week multi-university competition in which student teams build LLM-powered bots to influence a simulated election. We report on our custom social-media platform, the competition environment and design of its 4,000 AI-driven Non-Player Character (NPC) citizens, and what running Capture the Narrative at scale actually involved. In our first iteration, 108 teams from 18 Australian universities produced 7,068,206 player-bot posts, approximately 60% of all platform content. We surveyed 256 students before and 83 after the competition to understand their perceptions of misinformation and the game itself and found that students did not become more confident at spotting bots, contrary to what inoculation theory predicts. Because engagement was rewarded, most teams prioritised high-volume posting over nuanced influence, mirroring real-world platform dynamics. We close with recommendations for educators considering similar interventions, and propose future improvements, such as including a blue-team defensive phase.
