> ## Content Index > Fetch the complete content index at: https://model-philosophy.com/llms.txt > Use this file to discover other available public pages before exploring further. # Atlas of Unfulfilled Warnings: Government Wouldn’t Pass the Human CAPTCHA - URL: https://model-philosophy.com/atlas-of-unfulfilled-warnings-government-wouldnt-pass-the-human-captcha/ - Published: 2026-09-09T15:00:58.000Z - Updated: 2026-09-09T15:00:57.000Z - Author: James Rickard - Tags: theme:atlas-unfulfilled-warnings, philosopher:hannah-arendt, philosopher:max-weber, theme:bureaucracy, theme:robodebt, theme:2016-census > This article is a supplimental essay to the [Atlas of Unfilfilled Warnings](https://model-philosophy.com/atlas-of-unfulfilled-warnings/). # **Government Wouldn’t Pass the Human CAPTCHA** ## I. Proving the Obvious There must be six or seven copies of my driver's licence sitting in government databases by now. I know that I have uploaded it often enough. Government offices run our country, yet when they interact with us they seem to know very little about what is happening. They know when we were born, where we live, whether we've paid our taxes, whether we own a car, whether we crossed a border three months ago. Sometimes they know things we don't remember telling them. Governments have become remarkably good at collecting information. And yet, the experience of dealing with them does not feel like talking to someone who knows who we are. Every interaction begins from the start. You write your name on another form. You confirm your date of birth. You repeat your address. Somewhere, another copy of your driver's licence is uploaded for review. It is not that it is difficult, just tedious and time-consuming. We have built systems that remember everything except the last conversation. I suspect most of us don't mind, at least not at first. Bureaucracy has always carried friction. There are queues because there are people. There are forms because governments need records. Most of us accept that. But it is difficult to tell if the form exists because the government needs information, or because the system itself needs feeding. Every answer creates another question. Every document requires another document to verify it. Every proof of identity demands a second proof that the first proof belongs to you. That is the strange system of modern bureaucracy. We are told these systems exist to make life simpler, safer, and fairer. Most of the people who built them probably believed exactly that. I think they still do. But somewhere between policy, procurement, legislation, contractors, security reviews, budget cuts, and another election cycle, the citizen slowly disappears from view. What remains is the process. ## II. What We Inherit Max Weber worried, before our digital landscape, that modern society would become increasingly organised around bureaucracy, not because bureaucrats were evil or governments uniquely incompetent, but because bureaucracy was extraordinarily good at solving one problem. It could make decisions predictable. Predictability sounds like a virtue. We want judges who apply the law consistently. We want taxes calculated the same way each year. We want permits assessed by the same rules. Bureaucracy promises fairness by removing personality from the decision. Weber was looking at something larger than government forms. In The Protestant Ethic and the Spirit of Capitalism[\[1\]](#fn1), he was interested in how an ethic that began in religious life had escaped into the ordinary economic world: "*The Puritan wanted to work in a calling; we are forced to do so. For when asceticism was carried out of monastic cells into everyday life, and began to dominate worldly morality, it did its part in building the tremendous cosmos of the modern economic order.*” The first person chooses the system. The next person inherits it. Government bureaucracy is full of things that were probably reasonable decisions when someone first made them. Identity checks protect access. Procurement rules protect public money. Audits make decisions accountable. Standardised forms stop one person receiving different treatment from another. We did not set out to build something hostile to the people using it. We built safeguards. Then we inherited them. Inheritance does not make these systems wrong. It does make it easier to stop asking why they are there. Bureaucracies follow procedures so well that they gradually replace the very questions they were meant to answer. The form becomes more important than the person completing it. Verification becomes more important than understanding. We stop asking whether the outcome is just, because the process was correctly followed. “*The sad truth is that most evil is done by people who never make up their minds to be good or evil.*” Writing about the trial of Adolf Eichmann[\[2\]](#fn2), Hannah Arendt observed that extraordinary harm can emerge without extraordinary hatred. What interests me here is the disappearance of judgment. Harm does not always require someone who intends to cause it. Sometimes it requires people who stop asking what their small part, of a much larger process, is doing. Nobody needed to sit in a room and decide that innocent people should suffer for innocent people to suffer. Nobody building the Australian Census website hoped it would fail on the night it mattered most. Nobody writing procurement requirements sits in a meeting asking how they might quietly ruin somebody's afternoon. They are trying to complete a project. Stay within budget. Meet legislation. Pass security review. Deliver before the election. The suffering appears in the future, assembled from hundreds of perfectly reasonable decisions. ## III. The Machine Passed Its Audit Bureaucratic failure rarely looks like failure from the inside. Ministers stand behind lecterns announcing new digital services that will modernise government and reduce waste. Projects finish. Contracts are signed. Budgets are approved. There are photographs. Press releases. Demonstrations where everything works as expected. Somewhere, someone cuts a ribbon in front of a room full of cameras. Then ordinary people begin using it. In August 2016, Australia attempted something that sounded entirely reasonable. Rather than completing the national census on paper, the country would submit it online. It was presented as the next step in a modern digital nation, faster, cheaper, and more efficient than the system it replaced. On census night, millions of Australians attempted to do exactly what they had been asked to do. The website collapsed. For hours, the country watched an empty loading screen while explanations slowly emerged. There had been denial-of-service attacks. The protections had been disabled. Capacity assumptions had proved optimistic. Procurement decisions became technical explanations, and technical explanations became political ones. Inquiries followed. Reports were written. Recommendations were made. The census eventually happened. The night passed into history. Almost nobody remembers the names of the people responsible. The Senate inquiry[\[3\]](#fn3) that followed was eventually titled *2016 Census: Issues of Trust[\[4\]](#fn4)*. The servers were restored within days. Parliament spent considerably longer examining what their failure had done to trust. A few years later, Robodebt followed a similar path. This time the software did not merely inconvenience people. It accused them. Income averaging was treated as evidence of fraud, and letters demanding repayment began arriving in mailboxes across the country[\[5\]](#fn5). Some recipients paid debts they never owed because proving the government wrong required documents they no longer possessed. Others spent months trying to find someone with the authority to correct what everyone agreed looked unusual but nobody could change. Again, inquiries followed. Again, reports[\[6\]](#fn6) were written. Again, responsibility seemed to dissolve as investigators climbed the organisational chart. Ministers sat above departments. Departments interpreted policy. Lawyers gave advice. Public servants designed processes. Software turned those processes into calculations. Letters turned the calculations into debts. Responsibility existed at every level. Finding the place where it finally stopped was considerably harder. These are often described as technical failures. I am not convinced they are. Technical failures are the point at which the usability broke, but the system was broken before it was used. The software inherits the institution that built it. Every procurement shortcut, every committee compromise, every budget reduction, every attempt to spread responsibility across another department becomes part of the architecture. Code simply makes the bureaucracy visible. It executes, without hesitation, the values that were quietly embedded long before the first line was written. ## IV. Government Wouldn't Pass the Human CAPTCHA Every now and then a website asks us to prove that we are human. The questions are ordinary. Click every square containing a bicycle. Identify the traffic lights. Read the distorted letters. Tick the box that says you are not a robot. The computer is not really interested in bicycles. It is looking for the small inconsistencies that belong to living people. Humanity, it turns out, is difficult to automate. Governments have become rather fond of asking the same question. Prove who you are. Prove where you live. Prove you qualify. Prove you did not receive the payment. Prove the letter is wrong. Prove the database made a mistake. Prove that you exist. Most of us do. We upload another document. We wait in another queue. We spend another afternoon explaining ourselves to systems that already know our names. It is tiring, but we tell ourselves there must be a reason for it. Governments carry enormous responsibility. Verification matters. Public money should be protected. Fairness requires consistency. I think most people accept that. But, what would happen if we reversed the test? Suppose the government had to prove that it was human. Not by recognising bicycles or traffic lights, but by demonstrating the things we usually expect from another person. Could it admit that it had made a mistake before someone forced it to? Could it recognise that a process had become more important than the person standing in front of it? Could it apologise without first commissioning an inquiry? Could it remember that the purpose of a system is not to complete forms but to help the people completing them? These are not technical questions. They are human ones. Perhaps that is why they are so difficult. When the Senate investigated the Census, it examined procurement, governance, contracts and ministerial oversight. When the Robodebt Royal Commission examined one of the greatest failures of Australian public administration, it traced decisions across departments, legislation, advice, software and ministers. In both cases the investigations revealed remarkable detail about how the systems functioned. What they could never quite locate was a single place where responsibility finally came to rest. After all of those pages, hearings, reports and recommendations, responsibility becomes strangely difficult to find. It exists everywhere through the system, which sometimes seems to mean that it exists nowhere at all. --- 1. Weber, Max. The Protestant Ethic and the Spirit of Capitalism. United Kingdom: Taylor & Francis, 2005\. [↩︎](#fnref1) 2. Arendt, Hannah. Eichmann in Jerusalem: A Report on the Banality of Evil. United Kingdom: Penguin Publishing Group, 2006\. [↩︎](#fnref2) 3. Original Link: [Review of the Events Surrounding the 2016 eCensus](https://parlinfo.aph.gov.au/parlInfo/search/display/display.w3p;query=Id:%22publications/tabledpapers/a41f4f25-a08e-49a7-9b5f-d2c8af94f5c5%22?ref=model-philosophy.com "Review of the Events Surrounding the 2016 eCensus") Snapshot: [Internet Archive](https://web.archive.org/web/20260903150052/https://parlinfo.aph.gov.au/parlInfo/search/display/display.w3p;query=Id:%22publications/tabledpapers/a41f4f25-a08e-49a7-9b5f-d2c8af94f5c5%22 "Review of the Events Surrounding the 2016 eCensus") [↩︎](#fnref3) 4. Original Link: [2016 Census : issues of trust / The Senate, Economics References Committee](https://catalogue.nla.gov.au/catalog/7331369?ref=model-philosophy.com "2016 Census : issues of trust") Snapshot: [Internet Archive](https://web.archive.org/web/20260908135358/https://catalogue.nla.gov.au/catalog/7331369 "2016 Census : issues of trust") [↩︎](#fnref4) 5. “*Robodebt was a crude and cruel mechanism, neither fair nor legal, and it made many people feel like criminals. In essence, people were traumatised on the off-chance they might owe money. It was a costly failure of public administration, in both human and economic terms*” [Final report of the Royal Commission into the Robodebt Scheme](https://ministers.pmc.gov.au/gallagher/2023/final-report-royal-commission-robodebt-scheme?ref=model-philosophy.com "Final report of the Royal Commission into the Robodebt Scheme") Snapshot: [Internet Archive](https://web.archive.org/web/20260908135549/https://ministers.pmc.gov.au/gallagher/2023/final-report-royal-commission-robodebt-scheme "Final report of the Royal Commission into the Robodebt Scheme") [↩︎](#fnref5) 6. Original Link: [The Royal Commission into the Robodebt Scheme](https://robodebt.royalcommission.gov.au/publications/report?ref=model-philosophy.com "The Royal Commission into the Robodebt Scheme") Snapshot: [Internet Archive](https://web.archive.org/web/20260908135908/https://robodebt.royalcommission.gov.au/publications/report "The Royal Commission into the Robodebt Scheme") [↩︎](#fnref6)