Three design recommendations for gender-responsive AI literacy programs¶

Strategy · All strategies
Evidence · 3 claims (3 for) · 3 studies (2 associational, 1 causal), q2–q3 · 3 of 3 report an effect size · 2 claims rest on one study

Description¶

The article closes with three recommendations for AI literacy program design grounded in its gender-disaggregated findings: "(1) leverage workshops to increase female STEM career interest; (2) incorporate gender-specific deepfake safety modules, particularly for male students; and (3) ensure that workshop evaluations monitor equity of impact by gender." The recommendations follow directly from the observed baseline gaps, the female-biased gains, and the male-skewed deepfake sharing behaviour.

Design Implications¶

Context¶

Requirements¶

• Workshop evaluations must collect gender-disaggregated pre/post data to monitor equity of impact

Constraints¶

• Evidence base is limited to two co-educational regional NSW schools, which the authors state limits generalisability

Target Learners¶

• Australian secondary students in Years 7, 8, and 10

Target Learning Goals¶

• AI literacy
• STEM career aspiration development
• deepfake safety and ethical synthetic-media behaviour

Claims¶

• Deepfake behaviours show gender patterns: males more likely to have shared deepfakes, females more likely to have reported them [+W]
• Male secondary students report significantly higher STEM career interest than female students across AI, computer science, and engineering at baseline [+W]
• Gender is associated with small differences in GenAI use, literacy, and monitoring, while educational level shows no significant differences [+W]

Related Strategies¶

• Develop gender-specific well-being strategies, especially literacy support for male students

Examples¶

-

Key Sources¶

• Jake Renzella, Christian Bergh, Natasha Banks, and Alexandra Vassar. (2026). Gender Differences in AI Literacy Workshop Outcomes and Deepfake Engagement. https://arxiv.org/abs/2606.14718