Ban surveillance in schools

Ban surveillance in schools is the civil-society demand that schools and universities be free from biometric identification systems, AI behavior monitoring tools, and associated surveillance technologies — a framing that has propagated across K-12, higher-education, and international contexts through three distinct but structurally related controversies: the deployment of facial recognition in US public schools (Lockport City School District, 2018–2020); the mass adoption of AI remote proctoring platforms during the COVID-19 pandemic (2020–2021); and the use of biometric facial recognition for cashless school meal payments in UK schools (North Ayrshire Council, 2021–2023). The framing's core argument is three-part. First, students are owed heightened privacy protections in educational settings because the power asymmetry between schools and students — surveillance the student cannot refuse, using data the student cannot audit or contest — makes the consent frameworks that justify adult surveillance in other contexts unavailable. Second, facial recognition and AI monitoring systems are racially biased and disproportionately harm students of color, LGBTQ students, students with disabilities, and students in immigrant families — producing misidentification at higher rates for darker-skinned students and enabling targeting that concentrates on already over-surveilled populations. Third, surveillance in schools creates chilling effects on learning, free expression, and association — suppressing the exercise of rights and freedoms that educational settings are supposed to develop. The framing is structurally linked to the broader ban biometric mass surveillance demand and to the #BanTheScan campaign against police facial recognition, but it turns on an argument — the educational relationship's distinctive power asymmetry and the developmental context of children — that neither of those framings fully inhabits.

The Lockport controversy and Fight for the Future's campaign

The framing's K-12 focal controversy began in Lockport, New York — a small city in Niagara County twenty miles east of the Canadian border — where the Lockport City School District became the first known US public school district to deploy facial recognition. In 2015, a security consultant approaching the district offered a free threat assessment and recommended the purchase of a camera system with facial recognition capability. Following the February 2018 Marjory Stoneman Douglas High School shooting in Parkland, Florida — a moment of heightened national school-security spending — Lockport moved to install Aegis, a facial recognition system developed by SN Technologies. The Aegis system went live across 300 cameras in eight Lockport schools in January 2020, making Lockport the largest and most visible test case for school facial recognition in the United States.

The New York Civil Liberties Union sued the New York State Education Department over its approval of the system. The NYCLU's Stefanie Coyle stated that "Subjecting 5-year-olds to this technology will not make anyone safer, and we can't allow invasive surveillance to become the norm in our public spaces." The lawsuit crystallised the movement's central argument: that school safety rationales do not justify the biometric surveillance of children who are legally compelled to attend school and have no meaningful recourse short of withdrawal from public education.

Fight for the Future converted the Lockport controversy into a national campaign. In January 2020, the organisation partnered with Students for Sensible Drug Policy to launch a college campus campaign — mobilising student government resolutions at George Washington University and DePaul University, and collecting a 40+-rights-group coalition letter calling on universities to ban facial recognition as a threat to student and faculty safety. In September 2020, aligned with the back-to-school period, Fight for the Future pivoted specifically to K-12 schools: over 1,000 parents signed an open letter calling on school boards to prohibit facial recognition, naming the unknown psychological effects of constant childhood surveillance, the data-breach exposure risks, and the discriminatory targeting of students of color, girls, and gender nonconforming students as the three grounds for prohibition.

New York legislators passed a bill banning facial recognition in schools in July 2020; Governor Andrew Cuomo signed it into law in December 2020 — the first US state prohibition on facial recognition in schools. Fight for the Future's campaign page used New York's action as the model for the national demand: school facial recognition is "biased, and more likely to misidentify students of color"; it enables surveillance of movements, facial expressions, bathroom visits, and social interactions that administrators use to target students for minor infractions; and it produces watch lists and false arrests through errors that fall disproportionately on minority students. The campaign framed school facial recognition as an "unethical experiment" on children — a framing that established the asymmetry argument the movement has carried since: the conditions under which school surveillance occurs (compelled attendance, institutional power, developmental vulnerability) are not those under which adult consent in public space can be presumed.

Remote proctoring — the higher-education dimension

The framing's higher-education dimension emerged from the COVID-19 pandemic's rapid shift to remote learning in spring 2020 and the accompanying mass adoption of AI exam surveillance platforms. Proctorio, Respondus Monitor, HonorLock, ExamSoft, and ProctorU were deployed across hundreds of universities, monitoring students via webcam, tracking gaze patterns, analysing keystroke behaviour, and generating AI-scored assessments of suspicious behaviour — all running in students' homes, on their personal devices, with access to device file systems, in contexts where the option of refusing was indistinguishable from the option of failing the course. Proctorio alone was proctoring more than five times its pre-pandemic volume by the end of 2020.

The Electronic Frontier Foundation documented the student revolt in its September 2020 piece "Students Are Pushing Back Against Proctoring Surveillance Apps": petitions at dozens of institutions, most catalysed by documented tool failures. Students with darker skin tones reported being unable to be recognised by webcam-based identity verification before exams could begin. Gaze-tracking systems flagged neurodivergent students who look away from the screen as suspicious and routed them for disciplinary review. Proctoring installations accessed personal device files in ways students did not understand they had authorised. The Electronic Privacy Information Center (EPIC) asked the DC Attorney General to investigate the privacy practices and "unfair and deceptive" business claims of Proctorio, Respondus, Examity, ProctorU, and HonorLock.

At the University of Massachusetts Lowell, 1,200 students signed a petition objecting to Respondus, leading the university to announce professors could not mandate its use. The University of Southern California announced it would drop Respondus Monitor entirely. The proctoring controversy extended the school surveillance framing into higher education and into domestic environments — reframing the harm not as government surveillance in public space but as corporate surveillance purchased by universities and delivered into students' bedrooms, on their personal devices, in contexts where enrollment effectively removed meaningful refusal. The "students cannot meaningfully consent" argument that Fight for the Future had developed around children's facial recognition in K-12 schools turned out to apply structurally to higher education as well.

The UK dimension — biometric school meals

The framing's UK dimension illustrates how the demand extends beyond security-justified surveillance to cover convenience-justified biometric collection. North Ayrshire Council in Scotland introduced facial recognition for cashless meal payments across nine of its schools in 2021: a system in which a still image of each pupil was taken at the lunch queue and matched to a biometric template to deduct payment from an online account. Big Brother Watch had documented similar deployments at other UK schools and reported that some institutions were charging children for a swipecard if they declined biometric scanning — converting the nominal opt-out into a paid penalty.

After data protection concerns were raised with the Information Commissioner's Office (ICO), North Ayrshire Council ceased processing. The ICO issued a letter to North Ayrshire Council in January 2023 finding that the Council was unlikely to have met UK GDPR requirements for a valid lawful basis when processing children's biometric data. The Council had relied on parental consent, but the consent statement was not specific; it did not present facial recognition as an option distinct from the cashless payment system generally; and it failed to obtain consent directly from pupils aged 12 and over, despite Scottish law presuming those pupils to be of sufficient maturity to consent for themselves.

The North Ayrshire case added a dimension the US K-12 and proctoring controversies had not foregrounded: that the school surveillance demand does not require a security rationale. A sufficiently convenient administrative application — faster lunch queues, simpler payment management — provides the institutional justification, and the harms of biometric data collection at an educational institution fall on children regardless of whether the system was deployed to prevent threats or to process meals.

The ACLU's "Digital Dystopia" — civil-society consolidation

The framing's civil-society consolidation document is the American Civil Liberties Union's October 2023 report "Digital Dystopia: The Danger in Buying What the EdTech Surveillance Industry is Selling". The report documented a $3.1 billion US school surveillance market — cameras, AI content monitoring, social-media surveillance tools, facial recognition, and behavioural prediction systems — whose vendors make "unsubstantiated, unverifiable" safety claims backed by no outcome evidence. The report found that surveillance cameras were present in 8 of 10 of the deadliest school shootings in the United States without preventing them — a direct rebuttal of the industry's central marketing premise.

"Digital Dystopia" gave the chilling-effects claim the framing names a surveyed baseline: 6 in 10 students report they do not feel comfortable expressing their true thoughts and feelings online when their activity is monitored. The report's focus-group findings established that school surveillance negatively affected students' interactions with school staff, their communications with peers, their choices about which clubs and organisations to join, and their overall sense of safety at school. The populations the report identified as bearing disproportionate harm — students of color, LGBTQ students, students with disabilities, low-income students, and students in immigrant families — are the same populations that the Algorithmic Justice League's audit work on AI bias and Fight for the Future's campaign messaging had named since 2020.

Why it has carried

The framing's durability across K-12 facial recognition, remote proctoring, and biometric school meals rests on a structural feature: it names a single normative principle — the educational relationship's distinctive power asymmetry and the developmental vulnerability of children — and applies it across superficially different use cases. Whether the surveillance is framed as school security (facial recognition to detect threats), academic integrity (proctoring to detect cheating), or administrative convenience (biometric payment to speed lunch queues), the same argument reaches each: students cannot meaningfully consent because refusing participation in school surveillance is not meaningfully different from refusing participation in school.

This categorical argument is structurally harder to rebut than a proportionality analysis. A proportionality argument can be answered by adjusting the specific deployment — more safeguards, a narrower data scope, an opt-out mechanism. The "educational relationship makes meaningful consent unavailable" argument survives those answers because the opt-out mechanism does not resolve the power imbalance: a student who opts out of biometric lunch payment and pays a penalty for a swipecard, or opts out of remote proctoring and must explain herself to her professor, is navigating a penalty for non-compliance rather than exercising a genuine right. Fight for the Future's campaign, the ACLU's "Digital Dystopia" report, and the EFF's proctoring documentation all operate on this structural argument — which is why they converge on the same demand across different surveillance use cases and jurisdictions.

The framing's relationship to the coded gaze argument is direct: the bias documentation that Algorithmic Justice League and allied researchers produced about facial recognition systems' differential misidentification rates is the evidentiary backbone of the school-surveillance demand. A school that deploys facial recognition in a predominantly Black or brown student population is deploying a system documented to misidentify darker-skinned faces at higher rates, in an institutional context where a misidentification can result in disciplinary action, a watch-list entry, or a false-arrest referral against a student who has no practical recourse. The combination of biased system and power-asymmetric setting is what makes the "ban, not regulate" logic compelling: the harm is not correctable by procedural reform within a surveillance architecture that remains in place.