Four Surveillance Technologies Creating Challenges for Education

Abstract

Abstract“Ubiquitous AI”—embodied in cloud computing web services, coupled with sensors in phones and the physical world—is becoming infrastructural to cultural practices. It creates a surveillance society. We review the capabilities of four core surveillance technologies, all making headway into universities and PreK-12 schools: (1) location tracking, (2) facial identification, (3) automated speech recognition, and (4) social media mining. We pose primary issues educational research should investigate on cultural practices with these technologies. We interweave three priority themes: (1) how these technologies are shaping human development and learning; (2) current algorithmic biases and access inequities; and (3) the need for learners’ critical consciousness concerning their data privacy. We close with calls to action—research, policy and law, and practice.

Unpacking the Ethics of Using AI in Primary and Secondary Education: A Systematic Literature Review

“…AIPSED can also benefit students' skills, knowledge and learning ability, although some authors merely list such possibilities in passing (Huang, 2023;Mohammed & Watson, 2019;Nemorin et al, 2023;Nguyen et al, 2022;Yan et al, 2023). Many authors who engage in more depth with such claims focus on specific skills, with Pea et al (2023) noting that AI's better performance in tasks like pattern recognition could support teaching reading and writing (although they remain sceptical about AI's overall teaching ability to teach as a whole), or Su & Yang (2023) claiming that suggestions from tools like ChatGPT could improve students' academic writing. Kasneci et al (2023) argue that LLMs can improve writing ability by correcting grammatical and syntactic errors or helping students master specific writing styles (e.g., physics or literature), while students' reading comprehension could be increased thanks to the ability of such tools to summarise and explain difficult texts.…”

“…Furthermore, Pinkwart (2016) notes that privacy regulations differ between countries and that restrictions on data collection and use disincentivise development of AIPSED. To address some privacy challenges, Pea et al (2023) suggest that regular privacy checkups should be organised at schools to assess potential risks.…”

“…Some consider data collection as a form of surveillance (Akgun & Greenhow, 2022;Baker et al, 2023;Berendt et al, 2020;Farrow, 2023;Filgueiras, 2023;Gillani et al, 2023;Huang et al, 2023;Leaton Gray, 2019;Li & Gu, 2023;Mavrikis et al, 2021;Nemorin et al, 2023;Nugyen et al, 2023;Schiff, 2022). This could predominantly affect marginalised students (Bartoletti, 2022;Dieterle et al, 2022;Madaio et al, 2022;Pea et al, 2023;Selwyn, 2022), although Reiss (2021) and Selwyn (2022) stress that teachers might also be excessively monitored.…”

“…This theme synthesises the discussion of how and why AIPSED algorithms can lead to unfair, biased and inaccurate outcomes/decisions. Concerns about algorithmic fairness and bias (and the resulting discrimination) are raised by many authors (Akgun & Greenhow, 2022;Berendt et al, 2020;Bu, 2022;Corbeil & Corbeil, 2021;Crompton et al, 2022;Farrow, 2023;Filgueiras, 2023;Huang, 2023;Kasneci et al, 2023;Kitto & Knight, 2019;Lameras & Arnab, 2021;Leaton Gray, 2019;Li & Gu, 2023;Luan et al, 2020;Mavrikis et al, 2021;Nguyen et al, 2023;Pea et al, 2023;Saputra et al, 2023;Schiff, 2022;Su & Yang, 2023;Wei & Niemi, 2023). Specific concerns include a lack of attention to fairness in the development of AIPSED (Yan et al, 2023), excessive emphasis on certain pedagogies and learning styles (Schiff, 2021;Holstein & Doroudi, 2022;Porayska-Pomsta et al, 2023), prioritising developers' own cultures and worldviews (Mohammed & Watson, 2019;Schiff, 2021), presenting students with tasks poorly related with their ability that impede learning (Du Boulay, 2022a), difficulties surrounding the ongoing monitoring of the tools for bias (Dieterle et al, 2022), replication of historical biases present in educational practices and the society as a whole (Bartoletti, 2022;Holstein & Doroudi, 2022;Madaio et al, 2022;Rowe, 2019;Treviranus, 2022), and classification of students according to reductive categories (e.g., demog...…”

A Social Perspective on AI in the Higher Education System: A Semisystematic Literature Review

“…Surveillance. Continuous monitoring of student behaviors through AI can create an atmosphere of mistrust and potentially hinder academic freedom [38,127]. Shifting the education system to a surveillance culture can detract from the learning experience, making education feel more like a high-stakes monitoring environment rather than a space for exploration and growth [59,128].…”