This conference paper addresses gaps in scholarship and pedagogy surrounding the phenomenon of “dark patterns” in digital marketing and interface design by showcasing three curriculum-building projects. Dark patterns refer to a set of design strategies that co-opt the human-centred values advocated for in the fields of user experience (UX) design and human-computer interaction (HCI) to manipulate users into taking actions contrary to their personal interests. Recent dark patterns research has clustered within the fields of HCI, media studies, and game studies, with a focus on e-commerce and online gambling platforms. The presented projects put this established research into conversation with scholarship from business and marketing, science and technology studies, cognitive neuroscience, and disability studies to both create a more holistic definition of dark patterns and implement this expanded definition into university course curricula. These include Dark Patterns: Where Marketing Meets UX Design, focused on contextualizing dark patterns within historical market segmentation and merchandising strategies; Dark Patterns: Manipulative UX Design, on broadening the definition of dark patterns to include non-screen interfaces; and Designing for Normal and Failing Ethically, on analyzing how dark patterns have a disproportionate effect on individuals with certain cognitive disabilities. Collectively, these projects aimed to grant a greater historicity and social context to the phenomenon of dark patterns and introduce them as a utilizable pedagogical concept within the disciplines of communications, technology, and design. The findings of these projects are presented through the sharing of pedagogical materials, informal and formal feedback, and planned curriculum revisions.

As children grow up immersed in digital environments, scholars and policymakers emphasize the importance of helping children learn how to navigate privacy online. Prior work has found that educators recognize this need for privacy lessons but do not always feel equipped to teach them. Indeed, the term “privacy” has many meanings and the concept of privacy does not easily fit in a specific subject, intersecting with social studies, computer science, media literacy, digital literacy, and digital citizenship. Scholars have begun developing frameworks for privacy education, but such efforts will have a higher chance of success if they can be integrated into existing educational standards. Thus, in this study we are analyzing U.S. K-12 educational standards to understand whether and how they address privacy literacy. Our initial analysis has found that 44 of the 50 U.S. states have implemented educational standards related to privacy, largely as part of library, computer science, or social studies. The main privacy-related topics in state standards include being careful about posting information online and managing passwords. These preliminary findings suggest that while privacy is part of many state education standards, there are opportunities to help educators bring a more nuanced approach to privacy into their classrooms.

This paper draws on our research into the cultures of production surrounding the development of deepfakes and use of other forms of generative AI across public sites such as GitHub and YouTube, and subsequent reflective classroom experimentation and learning. Expanding on the notion of ‘vernacular pedagogies’ – informal and in situ education and relational literacy work – we propose a set of approaches for widening participation and involvement in AI and its underlying data practices. We reflect on the kinds of public vernacular pedagogy available on YouTube, GitHub and elsewhere online, and the kinds of experimental project work and learning environments that can be created with higher education students that can widen critical forms of AI participation and literacy.

Internet entrepreneurs, EdTech companies, AI enthusiasts, and other powerful stakeholders around the world have promoted the idea that big data and learning analytics (LA) have the potential to revolutionise education. LA, defined as the continuous measurement, collection, analysis and reporting of data about learners and their context (Gašević et al., 2015, p. 1), is increasingly being used to track and evaluate what students do in internet-mediated environments. A growing body of literature has questioned the benefits attributed to the use of AI-based solutions and raised a number of concerns about the current developments in the education sector. Despite this growing interest among researchers, we know little about how the beliefs, values and feelings of different groups of educational practitioners shape how they engage with AI-driven learning analytics technologies and influence the evolution of the cultures of practice shaping the adoption of learning analytics. In this paper, we report on research that asks: how do culturally situated beliefs, values and emotions shape practitioners’ engagements with narrow AI in different contexts of practice? The research project as a whole examines these cultures of practice across three contrasting contexts. Here we will discuss early findings from one of these contexts – learning analytics in higher education. With insights from this research, we aim to contribute to empower practitioners in higher education and relevant stakeholders to foster the development of critical and reflective data cultures that are able to exploit the possibilities of learning analytics while being critically responsive to their societal implications and limitations.