AI Workshops Boost Teens’ Ability to Spot Fake Videos and Images

Researchers are increasingly concerned about the potential harms of synthetic media, such as deepfakes, for young people, yet understanding of adolescent exposure and behaviour remains patchy. Christian Bergh, Alexandra Vassar, and Natasha Banks, from the University of New South Wales and Day of AI Australia, with Jessica Xu and Jake Renzella, investigated whether targeted workshops could improve Australian secondary students’ understanding of artificial intelligence, enhance their awareness of related safety issues, and potentially inspire greater interest in STEM careers. Their mixed-methods study, analysing data from 205 students before and 163 after an intervention, reveals a high baseline exposure to deepfakes, with over 80% having seen them, and demonstrates that the workshop successfully boosted students’ knowledge and ability to identify AI in everyday platforms. While the impact on long-term STEM aspirations was modest, these findings underscore the value of scalable AI literacy programmes in equipping young people with the critical skills needed to navigate an increasingly synthetic world.

Australian Secondary Students’ Initial AI Awareness and Post-Workshop Knowledge Gains were assessed via pre- and post-tests

Scientists have demonstrated a significant impact from workshop-based interventions designed to improve AI literacy among Australian secondary students. This research evaluates the effectiveness of Day of AI Australia’s curriculum in enhancing students’ understanding of artificial intelligence, its ethical implications, and its potential to influence STEM career pathways.
Employing a mixed-methods approach, the team analysed data from pre- and post-intervention surveys completed by 205 students initially, and 163 students post-intervention, to assess changes in knowledge, confidence, and career interests. Baseline data revealed a concerning level of exposure to synthetic media, with 82.4% of students reporting they had seen deepfakes, 18.5% sharing them, and 7.3% creating them.

The study unveils improved self-reported AI knowledge and confidence following the workshops, alongside a notable shift in how students perceive AI-driven systems. Prior to the intervention, students often viewed tools like Netflix, Spotify, and TikTok as simply “algorithm-based”, but afterwards demonstrated a greater understanding of their underlying AI foundations.

This suggests the workshops successfully fostered a more nuanced conceptual understanding of AI beyond surface-level interactions. Furthermore, the research establishes a positive correlation between workshop participation and increased interest in STEM careers, although the effect sizes were small, indicating the need for sustained educational efforts.

Researchers found that the intervention successfully broadened students’ recognition of AI in everyday recommendation systems, offering a practical connection between abstract concepts and their digital lives. The work opens possibilities for designing instruction that effectively links socio-technical AI concepts to adolescents’ lived experiences.

The findings support the implementation of scalable AI literacy programs that integrate foundational AI concepts with a specific focus on the safety aspects of synthetic media. This research contributes empirical evidence on the short-term impact of a workshop-based AI literacy intervention implemented in three Australian government secondary schools, providing a construct-level profile of students’ baseline AI literacy and ethical understanding.

Assessing changes in AI literacy, ethical understanding and STEM interest via pre- and post-workshop surveys is crucial for program evaluation

Researchers conducted a mixed-methods study evaluating the impact of Day of AI Australia’s workshop-based intervention on Australian secondary students in Years 7-10. The study employed pre- and post-intervention surveys, collecting baseline data from 205 students and post-intervention data from 163 students, to analyse changes in AI literacy and related behaviours.

Baseline data revealed that 82.4% of students had encountered deepfakes, with 18.5% sharing and 7.3% creating them, highlighting significant exposure to synthetic media. The research team administered surveys designed to assess students’ ability to identify AI in everyday tools, their understanding of AI ethics, training, and safety, and their interest in pursuing STEM-related careers.

Specifically, the pre- and post-intervention surveys contained questions gauging self-reported knowledge and confidence levels regarding AI concepts. Researchers analysed responses to determine shifts in understanding following the workshop. The study pioneered a method for measuring conceptual understanding of AI, moving beyond simple knowledge recall to assess students’ ability to recognise AI-driven systems rather than merely “algorithm-based” tools.

Scientists meticulously examined student responses concerning recognition of AI within popular platforms like Netflix, Spotify, and TikTok, establishing a direct link between workshop participation and improved identification skills. The team quantified changes in student perceptions using effect size estimates, focusing on key conceptual statements related to AI training, bias, and privacy.

Furthermore, the study assessed the impact of the workshop on students’ aspirations towards STEM careers, acknowledging that sustained interventions may be necessary to achieve substantial, long-term effects on career choices. This approach enables a nuanced understanding of the workshop’s influence on both cognitive understanding and attitudinal shifts among adolescent learners.

Enhanced AI awareness and knowledge following secondary student workshops proved beneficial

Scientists conducted a study evaluating an AI literacy workshop’s impact on Australian secondary students in Years 7-10. Researchers analysed changes in students’ AI knowledge, ethical understanding, and STEM career interest using pre- and post-intervention surveys with 205 students initially and 163 completing both surveys.

Baseline data revealed that 82.4% of students had seen deepfakes, with 18.5% sharing them and 7.3% creating them, highlighting significant exposure to synthetic media. Results demonstrate a measurable increase in self-reported AI knowledge and confidence following the workshop. Students exhibited improved recognition of AI within everyday platforms including Netflix, Spotify, and TikTok, indicating a shift in understanding from viewing these tools as simply “algorithm-based” to recognising them as AI-driven systems.

The team measured this conceptual shift through survey responses assessing students’ understanding of how these platforms function. Furthermore, the study recorded a post-workshop increase in students’ interest in STEM careers, although the observed effect sizes were small. Data shows that while the workshop sparked some interest, sustained educational approaches are likely needed to significantly influence long-term career aspirations.

Measurements confirm that the intervention successfully paired foundational AI concepts with an explicit focus on synthetic media safety. The research provides a construct-level profile of students’ baseline AI literacy, tool-use patterns, and ethical understanding. Effect size estimates of post-intervention shifts in self-reported AI knowledge, confidence, and conceptual statements regarding training, bias, and privacy were also calculated. This work supports scalable AI literacy programs designed to connect socio-technical AI concepts to adolescents’ digital lives and address the growing risks associated with synthetic media.

Enhanced AI literacy and perception shifts in Australian secondary school students are crucial for future success

Scientists have demonstrated that a workshop-based intervention can improve Australian secondary students’ understanding of artificial intelligence (AI) and its applications. The study assessed the impact of Day of AI Australia’s workshops on students in Years 7-10, utilising pre- and post-intervention surveys completed by 205 and 163 students respectively.

Analysis revealed a significant increase in self-reported AI knowledge and confidence following the workshops, alongside enhanced ability to identify AI within everyday platforms like Netflix, Spotify, and TikTok. Researchers found that 82.4% of students had previously encountered deepfakes, with a smaller proportion reporting creation or sharing of such content.

The intervention fostered a shift in student perception, moving away from viewing these tools as simply “algorithm-based” towards recognising them as AI-driven systems. While the workshops also showed a positive, though small, effect on students’ interest in STEM careers, the authors suggest that sustained educational approaches are needed to solidify these aspirations.

The study acknowledges a disproportionate distribution of Indigenous participants across schools, and a lack of specific analysis regarding their responses, representing a limitation. This research highlights the potential of scalable AI literacy programs to equip young people with the skills to navigate synthetic media safely and ethically.

The findings support integrating AI education more systematically into Australian curricula, potentially mitigating a widening digital skills gap. Future work should focus on longitudinal studies to assess the long-term retention of knowledge and the impact of sustained AI education on STEM participation, particularly prioritising culturally responsive frameworks to ensure equitable access for Indigenous students.