Community Voices
The people on the other side of the decision.
Most writing about automated decisions describes the system. These accounts start with the people it reached: students whose grades were set by a formula, teachers judged by a score they could not check, families flagged by a model none of them had heard of. In every case here, someone it reached spoke up: in court, in a complaint, in a march, or in writing.
Each account is drawn from the public record: court rulings, official inquiries, regulators, investigative reporting, and research. We summarise; we do not invent, combine, or dramatise. Every claim links to a source you can open and check.
What happened
The system, the decision it made, and who it reached, in plain words.
Who spoke up
How the people affected pushed back: a complaint, a lawsuit, a protest, testimony.
What changed
The ruling, reversal, settlement, or reform that followed, and what did not change.
Talk about it
Questions for a classroom or a kitchen table, with no single right answer.
Words worth sitting with
Lines from the public record and from Society & AI's own research, beside the question each account leaves. Select a card to read on.
- From our research Students and families become subject to decisions they cannot understand, contest, or meaningfully influence. The Rise of Algorithmic Societies: When Machines Govern Learning
- On the record “It's math that reflects our systems.” A Wisconsin education department spokesperson, defending the dropout warning system
- From our research A teacher might recognize that a student's recent performance decline stems from family instability, not academic incapacity. An algorithm sees only deviations from expected patterns. The Rise of Algorithmic Societies: When Machines Govern Learning
- From our research Systems that cannot be questioned cannot be trusted. What Is Society-Centered AI?
- On the record “face not found” … “room too dark” The exam software's messages to Robin Pocornie, by her account
- From our research Representation matters: those who bear the risks of a system should have voice in its creation. What Is Society-Centered AI?
- From our research The people who made the deployment decisions are not the people bearing the consequences. That gap, between who decides and who suffers, is a failure of causal accountability at civilizational scale. We Are Living in the Effects of a Cause We Never Authorized
- On the record “Ongekend onrecht.” Unprecedented injustice. Title of the Dutch parliamentary inquiry's report, 2020
- From our research The punishment for non-compliance is often invisible until it arrives. There is no trial, no appeal, no moment of confrontation with authority. Is AI Disciplining Society?
- From our research Systems fail silently, and no one is positioned to notice until the damage is done. The Human-in-the-Loop Economy
- On the record “a crude and cruel mechanism, neither fair nor legal.” Royal Commission into the Robodebt Scheme, 2023
- From our research A credible and humane assessment ecosystem requires a portfolio of measures, not a single test. Measuring Learning that Matters
- From our research Once people delegate a function to automation, they can become too willing to accept its recommendations or overlook contradictory information. AI Education Needs Authority Literacy Before Agent Literacy
- On the record No facial recognition lead in a lineup without “independent, reliable evidence.” Williams v. City of Detroit settlement, 2024
- From our research Accountability ties decisions to named owners, audit trails and remedies for harm. Public-Interest AI
- The question it leaves If a model is right about a school on average, is it fair to the student it is wrong about? The Summer an Algorithm Set the Grades
- The question it leaves When a tool cannot explain why it flagged your work, what should count as proof that you wrote it yourself? Flagged as AI-Written
- The question it leaves When someone says a system failed them, who should have to prove what, and with whose data? Face Not Found
- The question it leaves What does a label like "high risk" do to the student who carries it, even when nobody acts on it? Labelled High Risk in Ninth Grade
- The question it leaves If you cannot check how a score was calculated, can you ever really challenge it? A Score Nobody Could Check
- The question it leaves How much should one number, calculated once a year, decide about a person's work? From Effective to Ineffective in One Year
- The question it leaves When a system learns from public records, are the families who rely on public services the ones it sees most? A Risk Score on the Hotline
- The question it leaves When a model chooses whom to investigate, how do its choices turn into accusations? The Benefits Scandal That Brought Down a Government
- The question it leaves When an automated shortcut is wrong, who has to prove it: the government, or the person who opens the letter? Debts by Averaging
- The question it leaves A match is a guess. What has to happen before a guess becomes an arrest? A False Match, an Arrest at Home
Students
Grades, exams, and early warningsThe Summer an Algorithm Set the Grades
With exams cancelled, a model lowered 39 percent of teacher-assessed A-level grades, and small classes fared best. Students marched; the government reversed within days.
Exam grading model Read the account 2023 United StatesFlagged as AI-Written
An instructor asked a chatbot whether it wrote his students' essays. Researchers found detectors flagged most essays by non-native English writers. A label is not evidence.
AI-writing detection Read the account 2022–2023 NetherlandsFace Not Found
A student said proctoring software kept failing to detect her face. Her complaint to a human rights body ended in a split ruling worth reading closely.
Online exam proctoring Read the account 2012–2023 Wisconsin, USLabelled High Risk in Ninth Grade
A state system scored students on their chances of graduating, using race and income among its inputs. Most of its warnings were false alarms, and two students wrote about living with the label.
Dropout prediction Read the accountTeachers
Scores used to judge the people who teachA Score Nobody Could Check
Houston judged teachers with a score built on secret code that no one could verify. Teachers sued, and the district agreed not to fire anyone on an unverifiable score.
Teacher value-added scores Read the account 2016 New York, USFrom Effective to Ineffective in One Year
A teacher's state growth score dropped from 14 out of 20 to 1 in a single year. She challenged it in court, and a judge set the score aside as arbitrary.
Teacher growth scores Read the accountFamilies
Benefits, child welfare, and policing at homeA Risk Score on the Hotline
A county scores child-welfare calls using public records. Research found racial gaps in its recommendations, and parents with disabilities believe it flagged them.
Child welfare risk scoring Read the account 2021 NetherlandsThe Benefits Scandal That Brought Down a Government
A Dutch risk model that weighed citizenship helped wrongly brand tens of thousands of parents as fraudsters. Parents pushed back, and the government resigned.
Benefits fraud risk model Read the account 2015–2019 AustraliaDebts by Averaging
Australia averaged people's yearly income to decide who had been overpaid welfare. People with irregular pay were chased for debts they did not owe. A Royal Commission called it neither fair nor legal.
Automated welfare debts Read the account 2020–2024 Detroit, USA False Match, an Arrest at Home
Facial recognition wrongly matched a man's licence photo, and police arrested him in front of his daughters. His lawsuit changed how Detroit police may use the technology.
Police facial recognition Read the accountAdd your voice
Has an automated decision affected you?
We are gathering first-hand accounts from students, teachers, and families: a grade set by software, a flag you could not explain, an appeal that went nowhere, or one that worked. Tell us what happened in your own words. Nothing is published without your explicit agreement, you choose how you are named or whether you are named at all, we check the details with you before anything appears, and you can withdraw your account at any time. Use the contact form and mention Community Voices in your message.
Share an accountYours to give, and to take back
We publish an account only with the written agreement of the person it belongs to, or a parent or guardian for anyone under 18, and remove it on request.
Names are optional
Many people cannot speak openly about a school or an employer. We can publish without names, places, or any detail that would identify you.
Built for discussion
Each account ends with questions for a class or a family. They pair well with the Classroom Kits on explanation, fairness, and who decides.