INNOVAREModule 5 · Ethical AI

Case 1: The Kitchen Table

A letter. A government debt. A system that was certain it was right. Understanding why automated decisions without accountability produce the same failure, every time.

July 2026 · Case 1 of 6
As you read — hold this question

When an automated system makes a decision that affects a real person's life — and that decision is wrong — who is accountable for it?

380k
debt notices sent. $1.8 billion paid back. Three years the system ran before a Royal Commission found that the legal basis had never been properly checked.

The Australian government called it a data-matching initiative. Centrelink income data cross-referenced with Australian Tax Office records. Where a discrepancy appeared, a debt notice was issued. The method was income averaging: if your tax records showed annual income of $40,000, the system divided that by 26 and assumed you earned $1,538 every fortnight. For casual workers, seasonal employees, or people holding two jobs at different times, that average bore no relationship to what they actually earned in any given fortnight. The system couldn't distinguish between annual income and fortnightly income. It sent notices anyway. It was automated, scalable, and legally wrong from the start.

What Robodebt was
Quiz: Accountability

How income averaging produced 380,000 false debt notices

The Robodebt scheme (2016–2019) matched welfare payment records against Australian Tax Office data to identify discrepancies. The problem was the matching method. Annual ATO income figures were divided by 26 to produce a fortnightly average, then compared to Centrelink records. For anyone with variable income — casual workers, seasonal labour, people between jobs — the fortnightly average did not represent what they actually earned in any given period. The system treated a statistical artifact as a real debt.

Recipients who challenged the notices were required to produce evidence they did not owe money — reversing the normal burden of proof. Many could not locate years-old payslips or employer records. Many paid debts they did not owe because fighting a government system felt more dangerous than compliance.

The Legal Problem
The Royal Commission (2023) found that the income averaging method had no valid legal basis under the Social Security Act. Officers within Centrelink had flagged concerns before the scheme launched. The legal advice that would have identified the problem was never sought. The scheme ran for three years and generated $1.8 billion in recovered funds before it was ruled unlawful. The government repaid $1.8 billion. No one was prosecuted.
The pattern repeated

Five AI failures — different domains, same structure

Robodebt is not an outlier. The same structural failure — a system deployed without adequate accountability, producing harm at scale — recurs across domains and organisations. What changes is the setting.

System What it did The failure The accountability gap
Robodebt (Australia, 2016) Automated welfare debt detection using income averaging 380,000 invalid debt notices; no legal basis for the method No one checked the legal basis before deployment; those who flagged concerns were overridden
Microsoft Tay (2016) Public-facing conversational AI trained on Twitter interactions Produced racist and offensive content within 24 hours of launch No adversarial testing; no human oversight of live learning from public inputs
IBM Watson for Oncology (2017) AI cancer treatment recommendations deployed in hospitals globally Recommended treatments oncologists would not endorse; trained on single-hospital data MD Anderson invested $62M; no external validation of recommendations before deployment
COMPAS (US courts) Risk scoring for criminal reoffending used in bail and sentencing False-positive rates for Black defendants were roughly double those for white defendants at equivalent actual risk Algorithm was proprietary; defendants could not see or challenge the inputs used to score them
Air Canada chatbot (2024) Customer service AI handling bereavement fare enquiries Told a customer he could claim a bereavement discount retroactively — he could not Air Canada argued the chatbot was a separate legal entity responsible for its own actions; the tribunal rejected this
The accountability question

Whose name is on that decision?

When an AI system makes a decision that affects a real person — in a welfare system, a courtroom, a hospital, a customer service interaction — someone has to be responsible for that decision being right. The question is who. And right now, for most AI systems in the world, the answer is: nobody's.

Not "nobody was named." Nobody was designed in. The accountability gap is the default state when AI governance does not explicitly require someone to own each decision that the system produces.

The Air Canada case illustrates the end state of that logic. The airline attempted to disclaim the chatbot as a separate legal entity — arguing that the software's incorrect statement was the software's responsibility, not the organisation's. Tribunals have begun rejecting this framing. But the framing exists because organisations still treat AI as a tool rather than as an organisational decision-making agent that requires human accountability behind it.

Take this away

The accountability gap is not an accident. It is the predictable consequence of deploying AI without explicitly designing in human responsibility. If nobody's name is on the decision when it's made, nobody answers when it's wrong. Robodebt ran for three years. The same pattern will repeat until organisations treat accountability as a design requirement — not a post-incident assignment.

Quick recall — without looking back

Test yourself on this case

Question 1 of 3

What was the Robodebt income averaging method — and why did it produce invalid debt notices for variable-income workers?

Robodebt cross-referenced Centrelink welfare payment records with Australian Tax Office annual income data. When a discrepancy was found, the system generated a debt notice. The method divided annual ATO income by 26 to produce a fortnightly average, then compared that average to Centrelink records. For casual workers, seasonal employees, or anyone with variable income, the annual average bore no relationship to what they actually earned in any given fortnight. The system treated a statistical artifact — the fortnightly average — as a real debt amount. 380,000 people received notices. Many paid debts they did not owe. The Royal Commission found the income averaging method had no valid legal basis under the Social Security Act from the start.
Question 2 of 3

Describe the 'accountability gap' as illustrated by the Air Canada chatbot case. What does it reveal about how organisations currently assign responsibility for AI decisions?

Air Canada's chatbot incorrectly told a customer he could claim a bereavement fare discount retroactively. When challenged, Air Canada argued the chatbot was a separate legal entity responsible for its own statements — attempting to disclaim corporate accountability for what the AI had said. The tribunal rejected this. The case reveals how organisations attempt to use the autonomy of AI systems to shield themselves from accountability for decisions those systems make. The accountability gap occurs when no named human or role is explicitly responsible for the correctness of AI outputs — so when something goes wrong, there is no one to answer for it. Accountability must be designed in before deployment, not assigned after failure.
Question 3 of 3

What structural feature do Robodebt, Tay, Watson for Oncology, COMPAS, and the Air Canada chatbot share — and what does that pattern tell us about AI governance?

All five cases share the same structural failure: a system was deployed at scale without adequate accountability design, producing harm to real people before any governance mechanism caught the problem. In each case: someone assumed the system would work as intended without adversarial testing or human oversight of live outputs (Tay); or deployed without independent validation of the core methodology (Robodebt, Watson); or allowed the algorithm to operate without transparency or contestability for those it affected (COMPAS); or disclaimed organisational responsibility when the system produced incorrect outputs (Air Canada). The pattern tells us that AI governance cannot be reactive — it must be built into the design of what is deployed, who owns each decision, and who answers when the system is wrong.

Module 5 Videos

Module 5 · Short Video
Module 5 · Long Form · Whose Name Is On That Decision?

Sources

Royal Commission
Royal Commission into the Robodebt Scheme (2023). Final Report. Commonwealth of Australia.
Module content
BUSN9049 Module 5 — Ethical Considerations and Responsible AI. Flinders University, 2026.
Mikalef et al.
Mikalef, P. et al. (2022). Thinking responsibly about responsible AI and the dark side of AI. European Journal of Information Systems, 31(3), 257–268.
Air Canada
Moffatt v. Air Canada [2024] BCCRT 149. British Columbia Civil Resolution Tribunal.
Innovative Study
Long-form video: Whose Name Is On That Decision? Ethical AI Explained. Innovare Study, July 2026. youtu.be/tslQKmfxwk0