When a school introduces a platform, monitoring application, artificial intelligence or personal devices, it is not merely buying a tool: it changes who can teach, decide, observe and challenge a decision. Democratising school technology means enabling students, families, teachers and support staff to understand a proposal, influence it before it becomes irreversible and assess its effects afterwards. It does not require a vote on every update. It requires public questions, limited trials, clear accountability and a dignified route for anyone who is disadvantaged.
The everyday scene that exposes the problem#
Imagine that on Monday a family receives a notice: from the following week, every assignment must be submitted through a new platform. The school says it will make monitoring easier. Teachers have watched a thirty-minute demonstration; students were not involved; nobody asked how many households share a device, what data the supplier collects or how a student using a screen reader will work. When difficulties appear, they are described as resistance to change. This is not only a technical problem. A decision that reorganises daily life has been presented as a completed fact.
Imagine that on Monday a family receives a notice: from the following week, every assignment must be submitted through a new platform.
Technology concentrates many small choices with a large combined effect: which behaviour is recorded, what counts as participation, which alerts reach a family, what content a system recommends and what alternative a teacher retains. Although no single choice looks momentous, together they may reshape assessment, privacy and access. School democracy therefore cannot be limited to announcing a purchase afterwards. It must operate while needs can still be compared, options rejected and conditions established. Participation is not collecting decorative opinions; it means showing what changed because people spoke.
Begin with a need, not a product#
Before contacting suppliers, the team can describe the problem on one page. Is feedback arriving too late? Is accessible communication missing? Do learners need practice with a particular skill? It should then identify who is affected, what evidence confirms the difficulty and how people manage it today. This pause prevents an attractive feature from manufacturing its own necessity. It can also reveal organisational solutions: protected tutorial time, simpler channels or shared materials may work better than another account, another password and another stream of personal data.
The opening page should state boundaries. For example: facial recognition will not be used; an automated recommendation will never determine a final grade; an activity will remain possible without a personal phone; data will not support advertising; and an identifiable person will handle errors. Boundaries make bids easier to compare and protect a school from accepting conditions merely because staff have already invested time. If a tool works only by surrendering an essential safeguard, it does not fit, however impressive its demonstration.
Who needs a place at the table#
A useful decision group need not be enormous, but it must bring together different experiences. It might include a leader, digital coordinator, two teachers from different age groups, special educational needs or counselling staff, administration, family representatives and students old enough to participate. It should deliberately include people who often pay the hidden cost: a household with limited connectivity, someone who uses assistive technology, or the teacher who will later solve incidents. If everyone around the table owns a recent device, has reliable broadband and feels digitally confident, important risks will remain invisible.
Students possess information that is absent from a contract. They know when notifications create pressure, when several teachers demand the same task in different places, and which shortcuts emerge to survive a nonsensical procedure. Participation should match age: guided journeys for younger children, scenario workshops in secondary school and formal representation for high-impact decisions. Asking does not require accepting every preference. It requires giving reasons and distinguishing convenience, safety, pedagogical purpose and rights.
Families are not a single homogeneous group either. One online survey can exclude precisely those facing the greatest barriers. A short form should be combined with face-to-face or telephone conversations and translated or accessible material. Questions need to be concrete: which device would the child use, what new cost or time arises, what information would help a parent without turning them into a monitor, and what alternative would be workable? Publishing an anonymous summary of answers and decisions closes the loop. Without feedback, consultation becomes an extraction of opinions.
A public record for every tool#
The community needs a readable record, not one hundred pages of legal terms. It should state the educational purpose, year groups affected, data collected, people with access, supplier, retention period, hosting location where relevant, automated decisions, total cost and incident channel. It must also explain what happens if a family declines an optional use or if connectivity fails. The alternative has to enable the same learning without penalty and without publicly identifying the person who asks for it.
Transparency includes acknowledging uncertainty. If the school does not yet know how often an automated classification fails, it can say so and make that a trial question. Pretending certainty prevents learning. Where artificial intelligence is involved, the record should explain which tasks permit it, what human check exists and which materials must not be entered. Writing “AI is used” is insufficient. A tool that drafts practice questions presents different risks from one that detects emotions, predicts dropout or accuses a learner of cheating.
Test on a small scale, measure well and retain the power to stop#
A democratic trial begins small: one group, one purpose, four or six weeks, and no irreversible academic consequence. Before it starts, the school records its baseline. Measures might include teacher time, actual access, missing assignments, support requests, feedback quality and privacy incidents. Short accounts from students and families matter too. The purpose is not to prove that a purchase was wise. It is to discover the conditions under which the tool contributes value and the point at which it should stop.
The protocol should define stopping rules before enthusiasm takes over. Pause the trial if it repeatedly excludes a group, if an important result cannot be explained, if previously undisclosed uses of data appear, if the alternative does not work, or if incidents consume more resources than expected. Ending a trial is not failure; it is using reversibility to prevent greater harm. The final decision may be to adopt, modify, run a different trial or withdraw. Each is legitimate when supported by evidence.
Evaluation should separate averages from distribution. A better average submission time does not compensate for one identifiable group falling consistently behind. Results can be compared by age, support need, device availability and language without publishing data that identifies individuals. Displaced work must also be counted. If a teacher saves ten minutes while a parent spends an hour recovering passwords, the system has not become more efficient; it has transferred cost to someone with less bargaining power.
Procurement, contracts and dependency#
Participation loses force if it arrives after a long contract is signed. Democratic requirements belong in procurement: export of student work in usable formats, verifiable deletion, accessibility, support, information about subprocessors, notice of changes and an exit route that preserves the educational record. Interoperability matters as well. A school dependent on one supplier may retain a mediocre product because migration is too expensive. Planning the exit before entry protects both the budget and pedagogical autonomy.
Not everything can be settled locally: legal, curricular and security obligations apply. Democracy does not mean that a majority may remove a minority's rights or authorise unnecessary processing of personal data. Leaders should explain which aspects are fixed by regulation and where the school has discretion. Questions about privacy, accessibility or contracts should go to the responsible professionals, and the answer should be documented. A school council deliberates about purposes and conditions; it does not replace legal or technical assessment.
A right to review that works in practice#
Any system that classifies, recommends or sends alerts needs human review. Students should know that they can ask what happened without being labelled difficult. A channel may begin with the teacher and escalate to an accountable person within a stated time. The reply must provide criteria and evidence, not “the platform says so”. When an error reveals a pattern, the process must be corrected too: review settings, notify other affected people and record the improvement. Quietly fixing each individual case allows the same fault to continue.
A simple decision log creates institutional memory. For each tool it records date, need, participants, alternatives considered, trial, indicators, incidents, decision and next review. It should contain no intimate student details. A new team can understand why something was chosen, and a temporary solution is less likely to become permanent through neglect. An annual review might confirm use, restrict features or close the service. The question is not whether the tool still operates, but whether it remains necessary, proportionate and educationally worthwhile.
Questions a family can take to school#
A family can ask: “What specific problem does this solve?”, “What does my child lose without a smartphone?”, “Who can see their information?”, “How is an error reviewed?” and “When will you decide whether it continues?” No technical vocabulary is required. If answers remain vague, parents can request the public record and raise the issue through family representatives or the school council. A collaborative approach is usually most effective: describe an observed situation and suggest a feasible improvement. Criticism becomes useful when it links a principle to verifiable experience.
At home, adults can ask a teenager what helps, what distracts, and what the platform seems to misunderstand. Listening does not mean endorsing every suspicion; it gathers information before speaking to school. If there is immediate harm —a grade, data exposure or an access barrier— keep screenshots and dates without circulating information about other children. Then request review through the agreed channel. Responsible participation combines voice, evidence and care.
A workable plan for the next term#
In week one, the school identifies one need and publishes the initial record. In week two, it assembles the diverse group and gathers input through several channels. In week three, it compares alternatives, including buying nothing, and sets boundaries. Weeks four to eight hold the supported trial and incident log. In week nine, the group examines disaggregated results and lived experiences. In week ten, the responsible body publishes a reasoned decision, changes made and the next review date. The timetable makes it visible that participation can alter an outcome.
UNESCO's Global Education Monitoring Report on technology identifies access, governance and teacher preparation as conditions for technology to contribute value. The European Commission's Digital Education Action Plan calls for high-quality, inclusive and accessible digital education. The Council of Europe approaches digital transformation in education through rights and citizenship, while the Spanish Data Protection Agency's guide for educational centres helps schools translate responsibility for personal data into practice.
A democratic school neither fears technology nor idealises it. It chooses the ends a tool serves, limits its power, listens to those who bear its consequences and retains the ability to change course. A simple test asks who could speak beforehand, what information they received, what changed because they participated and how they can leave if something goes wrong. When those answers are visible, innovation stops being an order delivered from above and becomes collective learning. This way of governing is itself an education: it shows children and teenagers that a digital future is not merely consumed. It is debated, cared for and built.
