AI, Deepfakes, Psychosocial Risk and the National Higher Education Code
Technology-facilitated gender-based violence is becoming an emerging governance challenge for higher education. This research walkthrough examines deepfakes, image-based abuse, smart glasses, covert recording, doxing and psychosocial risk through the National Higher Education Code to Prevent and Respond to Gender-Based Violence and its Regulatory Guidance. Link to the Guidelines here.
This walkthrough is intended for university leaders, researchers, policy and governance professionals, gender-based violence prevention teams, WHS and psychosocial safety practitioners, educators and others considering how emerging technologies intersect with the National Higher Education Code.
Please cite as: Arantes, J. (n.d.).Technology Facilitated Abuse in Education. https://www.janinearantes.com/
What changes when deepfakes, image-based abuse and other digitally mediated harms are understood explicitly as forms of gender-based violence?
This is where my thinking starts. I have researched cyber abuse, deepfakes, digital identities and technology-mediated harm for several years, but the Regulatory Guidance gives these previously connected concerns a much sharper regulatory location, because “technology-facilitated abuse (including image-based abuse)” (Regulatory Guidance, p. 14) is part of the plan. It made me return to my earlier work and ask what becomes visible when these harms are understood explicitly as gender-based violence. What follows is a meander through old posts to bring it altogether around the new Code.
A short provocation that shifts the deepfake conversation beyond misinformation and scams to gendered harm, women’s safety, workplaces and anticipatory governance in education.
This post considers image-based abuse in schools from the perspective of both students and educators, drawing attention to the gendered nature of deepfake material and the need to include teachers in institutional conversations about rights, prevention and safety.
Emerging Technologies and Duty of Care: What does a safe educational environment mean when emerging technologies can create new pathways for surveillance, capture, manipulation and harm?
This is where my thinking starts to move from the technology itself to the environment into which it is introduced. Smart glasses, deepfakes and covert recording have led me repeatedly to questions about whose safety is considered when technologies enter classrooms and campuses. The Regulatory Guidance states that “duty of care is not limited to physical safety but includes psychological, cultural and emotional safety” (p. 20). For me, that opens a much larger question about what a safe educational environment now means when technological capabilities themselves can create new conditions for harm.
This post explores what changes when image capture becomes less visible. It connects smart glasses with deepfakes, covert recording, gendered harms and the possibility that an ordinary educational interaction can rapidly become manipulated digital content.
This piece examines covert recording through wearable technologies and considers what these technologies mean for consent, privacy, sensitive institutional activities and the safety of staff and students.
This article considers surveillance, recording, data security and educational policy as wearable AI enters learning environments. It provides an accessible entry into the governance questions raised by technologies that can record without the familiar signals associated with phones.
Who is responsible for anticipating technology-facilitated harm before a technology becomes embedded in institutional practice?
At this point my thinking moves upstream. If emerging technologies can alter the conditions through which harm becomes possible, responsibility cannot sit solely with individual person, lecturers or students. I keep returning to questions of who anticipated the risk, who authorised adoption, what was assessed, and where accountability sits. The Regulatory Guidance describes prevention and response as “a core strategic and operational priority” (p. 35). That pushes my work on AI governance toward a more explicit consideration of institutional responsibility for technology-facilitated harm.
This paper brings smart glasses, covert livestreaming, doxing and deepfakes together as institutional governance issues. It introduces the SHIELDS Framework, combining privacy and psychosocial risk assessment to support anticipatory decision-making around emerging technology.
This post translates the working paper into the institutional context, connecting wearable technologies with privacy, psychosocial risk, accessibility, duty of care and the National Higher Education Code.
What do students, staff and leaders need to understand about AI beyond how to use it?
This section makes me think differently about AI literacy. Much of the education conversation has concentrated on knowing how to use AI, detect it or teach with it. I am increasingly interested in whether our knowledge and capability extend to recognising what these technologies can enable around power, gender and harm. The Regulatory Guidance states: “This education and training should go beyond awareness-raising; it should promote behaviour change by encouraging critical reflection on gender, power and violence” (p. 26). That sentence substantially widens what I think technology education in higher education now needs to encompass, and prompted the Smart Glasses on Campus work (amongst others below).
This piece broadens AI literacy beyond assessment and academic integrity to include deepfake abuse, doxing, privacy breaches and psychosocial risk, particularly for women, LGBTIQ+ students and students experiencing marginalisation.
This post connects GenAI with the realities of a feminised education workforce, arguing that smart glasses, deepfakes and covert surveillance also need to be understood as gendered workplace-safety concerns.
Privacy, Consent and Institutional Safety: What does meaningful consent look like when technologies can record, retain, reproduce and manipulate people with increasing invisibility?
Privacy has been running through my research since my earliest work on educational data, but wearable AI and generative technologies have changed the problem. A face, voice or interaction can now be captured, retained, manipulated and redistributed in ways that are difficult to anticipate at the point of collection. The Regulatory Guidance is unequivocal that “Privacy is a fundamental human right” (p. 27). I read that alongside my smart glasses and deepfake work and find myself asking whether our existing ideas of consent remain adequate for these emerging technological conditions.
When AI and tools like smart glasses are present on campus institutions need to balance privacy, safety, transparency and their responsibility to prevent and respond to gender-based violence. The smart-glasses resources above are particularly relevant here because they expose an emerging governance problem: increasingly unobtrusive technologies can capture identifiable images, voices, behaviours and contextual information without the obvious social signal associated with conventional recording.
https://www.linkedin.com/posts/janine-arantes_aige-activity-7381548417887543296-pJEM
https://www.linkedin.com/posts/janine-arantes_aige-smartglasses-activity-7407877399155085312-VQ3f
https://www.nature.com/articles/s41599-026-06988-5
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5495783
Intersectionality, Visibility and Data: Who becomes more exposed to technological harm, who becomes less visible, and how are those differences shaped by existing inequalities?
This takes me back to some of my earlier work on LGBTQI+ data, algorithmic exclusion and who becomes visible or invisible within technological systems. Technology-facilitated harm cannot be separated from the identities, inequalities and power relations already present within institutions. The Regulatory Guidance recognises that “experiences of violence are shaped by overlapping identities, power dynamics and systemic barriers” (p. 19). That makes my earlier data justice work feel less peripheral to this discussion and more foundational to understanding who may experience emerging forms of harm differently.
This work examines what happens when LGBTQI+ lives become absent or inadequately represented within data systems. It provides a critical foundation for thinking about visibility, identity, automated decision-making and intersectional governance.
This article translates the wider data-justice argument into public policy, asking what happens when particular identities and experiences are missing from the datasets through which institutions understand populations and make decisions.
How do institutions govern emerging risks when technological capability is developing faster than the evidence, policy and regulatory responses around it?
This is where the page deliberately remains unfinished as of August 2026.
Emerging technologies are changing quickly, and I do not think the research questions around deepfakes, wearable AI, synthetic media and technology-facilitated abuse are settled. The Regulatory Guidance asks institutions to “leverage existing expertise within the institution, including researchers and educators working on gender equality, social inclusion and violence prevention” (p. 25). I see this collection as part of that continuing work: revisiting earlier research, testing new questions and adding evidence as the technological and regulatory landscape changes.
Watch this space!
Discussion Questions
What is technology-facilitated gender-based violence in higher education?
Does the National Higher Education Code include technology-facilitated abuse?
How do deepfakes intersect with gender-based violence in universities?
What do smart glasses mean for privacy and consent in higher education?
How does psychosocial risk intersect with the National Higher Education Code?