2000, George W. Bush visits an elementary school computer lab in Oregon. Surveillance of private citizens in the U.S. became increasingly and normalized in the 21st century.
With all the buzz on AI being implemented in higher education, most don’t realize that the use of AI is being facilitated by the large datasets that colleges have collected on their populations in the 21st century. Any type of data from wifi connections, to football scores, to grades, to library downloads are ripe for AI training.
The Surveilled Student
By the 2000s new technologies that could efficiently gather and utilize student data to develop predictive models of their behavior began to infiltrate education:
Electronic ID cards could track student movement and purchases
Wi-fi connections to track students
Online learning systems like Blackboard to track student performance
Email opens & clicks to monitor online engagement
Such data can be used to develop algorithms for aggregating student data & making predictions (Gardner 2019; Rampell 2008; Foote 2019). Often students were unaware of such monitoring. The data was used in predictive models based on incoming college stats were used to identify at-risk students
Purdue University was one of the first universities to track students using an in-house platform called “Signals” (Gardner 2019). The program worked to identify at-risk students by monitoring their GPA, standardized tests scores, dining hall attendance, financial aid status, and logging in use to monitor their risk of failing and used common traffic light signals to signal their academic status (Rampell 2008). The implementation of learning management systems like Blackboard were ways to gain insight into student data. Sounds useful right?
However, students in the introductory biology course were not aware that they were being studied as part of the Signals program. An associate biology professor Laurie E. Iten at Purdue was quoted by The Chronicle of Higher Education about the Signals program “I don’t care what they think of the program; I only care if it improves their grades,” (Rampell 2008).
A follow up study in 2016 suggests that Signals was an inspiration for learning analytics at other institutions but was not without controversy. While, there were 10% more As and Bs awarded in courses using Signals than without (CASE STUDY A: Traffic lights and interventions: Signals at Purdue University 2016), there were concerns about protecting teaching and student performance privacy, and two minority students reported being demoralized by messages sent from Signals.
Critics of surveillance point out the unethical implications of this data collection in that it violates data security and privacy. While many say that students “don’t care” about this data collection, there tends to be a gap in students understanding about how the data is being safeguarded and used and their opinions on data collection. Some critics say that this data monitoring is a type of “grooming” for creating a more compliant generation. Sophia Cop, a senior staff attorney at the Electronic Frontier Foundation, which is a nonprofit focused on civil liberties and technology, is quoted by the Chronicle of Higher Education that “We are grooming the younger generation to become used to being tracked and being surveilled and not being bothered by it” (Gardner 2019).
Additionally, quantitative metrics discount and individual student’s personality and can be used to stereotype a student. A controversial Washington Post opinion article ominously entitled “Someone is Watching You” by Purdue’s president Mitchell E. Daniels (Daniels 2018) epitomized the divide between student and administration’s view of student privacy. In the article, Daniels writes in an upbeat tone about the “treasure trove” of data that Purdue is collecting on its students (Daniels 2018). He writes “We know where each student is anytime — which is virtually all the time — their mobile devices are connected to our WiFi network” (Daniels 2018).
Data to replace human teachers??
According to Marc Bousquet’s 2008 book “How the University Work”, the “informationalized” university attempts to convert any form of labor, paid or unpaid, into a surplus value, from custodial work, to editing scholarly journals, to score keeping at games. Such data can be analyzed to make predictions about student, teacher, staff, and donor behavior (high football scores might mean more donations for example).
Once these processes are thoroughly understood by the university, typically through the deployment of technological surveillance, labor becomes “flexible” (62). Not only can faculty be surveilled through technological systems, but their workload increases due to the time required for learning these technological systems and by allowing for continuous access via email and online calendars.
These technological systems quietly gather and store data on their classes (i.e., lesson plans, assignments, syllabi, student evaluations) Then, tenured faculty can be displaced by graduate students, part-time teachers, technology specialists, writing consultants, and other lowly paid, temporary positions. The goal was not to create a completely “digital” university, but to retain the student experience and replace the staff with cheaper instructional options.
“Informationalized” labor takes the load off the institution and displaces it to social institutions (64). Education, once a public good, has been put into private hands with the piloting of educational AI technologies in late 2010s.
AI has been used in online chats boxes on school websites, to send automated email responses, in plagiarism detection software, and in grammar and spelling apps. AI has been used more controversially in grading, reaccommodating classes, and evaluating admissions decisions (Newton 2021). Many argue that this can relieve the increasing workload that technology has placed on faculty. However, unfortunately, AI algorithms can perpetuate bias and can be an invasion of privacy.
Putting Data to Work
With the advent of AI, and increasing university partnership of AI, it is concerning because these universities have collected huge datasets of syllabi, exams, notes, recorded lectures, and evaluations about classes. This data collection has been likely going on for at least 2 decades.
It is not a coincidence that since the 2008 financial crises, the number of adjunct instructors has increased in American universities while tenure-track job openings have remained fairly stagnant. Bousquet’s 2008 warnings were correct. Indeed in 2025, we are seeing AI-generated textbooks from previous course materials.
Cover of “History & Fiction: Survey of Literature from the Middle Ages to the 17th Century”, the UCLA AI-generated textbook. According to James Folta at Lithub, the images and words are nonsensically organized.
Disturbingly, an op-ed in Inside Higher Education entitled “Faculty Must Protect Their Labor from AI Replacement” by John Warner writes about the infamous comparative lit textbook generated by a professor’s notes at UCLA:
The professor argues—I would say rationalizes—that this is good for students because “Normally, I would spend lectures contextualizing the material and using visuals to demonstrate the content. But now all of that is in the textbook we generated, and I can actually work with students to read the primary sources and walk them through what it means to analyze and think critically.” […]
I actually find it shocking that anyone would give over their intellectual property for such an exercise, which makes me wonder if this professor is being compensated beyond their base salary for pursuing such “innovations.” If not, it’s an absolutely foolhardy choice. If so, it’s selling out all future instructional laborers for individual gain.
Either way, it’s the pattern of adjunctification repeating, as relatively well-off tenured faculty protect their individual privileges by permitting the future immiseration of others.
Ward offers good news that at least a faculty member at the Universities of Wisconsin System pushed back against changes in copyright:
“they believe would cheapen the relationship between students and their professors and potentially allow artificial intelligence bots to replace faculty members.”
In essence, the institutions are claiming copyrights, “a non-exclusive license to use syllabi in furtherance of its business needs and mission,” over faculty instructional materials, the very things that are being used in the above example of the comparative lit course at UCLA.
This brings up very pertinent questions of copyright. Who really “owns” the syllabus a professor carefully crafted for their class?
Big Brother is Always Watching?
Generation Z (defined as being born in 1997-2012) and younger Millennial students (defined as being born between 1981-1996) have grown up in a culture of surveillance. The passage of the 2001 Patriot Act during the Bush administration sanctioned mass surveillance in the name of safety from terrorism. In relation to school, surveillance can be justified for identifying campus threats, such as identifying potential mass shooters or students with an intention to self-harm (Beckett 2019; Gardner 2019).
However, these algorithms can target people of different cultural back grounds, disabilities, sexual identities, or students of color. Students’ private emails, Internet searches, and social media accounts are sources which are mined for such information. These companies then send such flags to the local school. The ACLU condemns such actions as solutions for preventing threats and argues for better mental health services or anti-bullying programs. The ACLU argues that surveillance constricts students’ intellectual freedom, freedom of speech, freedom of association, expectations for privacy, and result in false identifications which can cause psychological harm to students and increase racial biases (Marlow 2021; Nance 2017). Surveillance also provides no deterrence to gun crimes.
Like the so-called “Patriot Act”, often such measures are used to justify keeping students safe, when often, they provide little to no deterrence. In the 21st century, the role that University played in protecting students and regulating student behavior came into question again.
Strict social codes and monitoring were the norm pre-1960/1970s at most universities, and many undergraduate students had curfews, strict dress codes, and strict moral codes—in a way colleges assumed certain responsibilities with student’s parents.
1945, female students protest ‘no pants’ rule.
In an era of increasing concerns about legal liability the return of in loco parentis may be upon us. Highly publicized student deaths due to hazing and partying have increased University oversight over student organizations. Additionally, highly publicized student deaths, an increasing mental health crisis among students (Bernstein 2007; Patel 2019), and the H1N1 and COVID-19 pandemics increased concern about the role that Universities play in regulating student life.
Conclusion
In all, there is no easy answer, and no way to turn back the clock on these data innovations. While academia is one of the most unionized workforces in America, it has only been marginally successful at stemming back the adjunctification of universities. These large waves of data technological development are flooding through STEM an humanities fields (Furthermore, UCLA now offers a class called “Algo-Lit” that studies AI-generated literature).
Bernstein, Elizabeth. 2007. “Bucking Privacy Concerns, Cornell Acts as Watchdog.” WSJ. https://www.wsj.com/articles/SB119881134406054777 (April 27, 2021)
Daniels, Mitch. 2018. “Opinion : Someone Is Watching You.” The Washington Post. https://www.washingtonpost.com/opinions/its-okay-to-be-paranoid-someone-is-watching-you/2018/03/27/1a161d4c-2327-11e8-86f6-54bfff693d2b_story.html?noredirect=on
Foote, Keith D. 2019. “A Brief History of Machine Learning`.” DATAVERSITY. https://www.dataversity.net/a-brief-history-of-machine-learning/#
Gardner, Lee. 2019. “Students under Surveillance.” The Chronicle of Higher Education. AX
Marlow, Chad. 2021. “Student Surveillance Versus Gun Control : The School Safety Discussion We Aren’ t Having.” ACLU
Nance, Jason P. 2017. “Student Surveillance, Racial Inequalities, and Implicit Racial Bias.” Emory Law Journal 66(4): 765–837.
Good read, thanks for sharing!
Do you think there are any good types of data for universities to collect and better understand students? In other words, is there a “right” amount of data to harvest? Or is it all bad?