← Case Library

Robodebt (Australian Government)

Q3
Government · Spear · The Trust Breach
Researched as of October 2026
ShareLinkedInX
AdaptAugmentAssist
Q3Automator
Q6Optimiser
Q9Conductor
Q2Tactician
Q5Partner
Q8Explorer
Q1Operator
Q4Weaver
Q7Navigator
StructuredDynamicComplex
2/9 cells occupied
Shape Read

The Royal Commission records that averaged PAYG data was applied automatically where alternative information was not provided and accepted, and that the aim was to reduce and preferably remove human involvement. In the staged rollout to 30 September 2016, ninety-five percent of reviews were completed with no action by a compliance officer. Robodebt automated the calculation and recovery of welfare debts by averaging annual tax data into fortnightly income figures, a Q3 Automator execution layer applied to a task the Social Security Act structurally required to be Q2 Tactician case-by-case human judgement. The system was technically functional at industrial scale, issuing up to 20,000 debt notices a week and recovering hundreds of millions of dollars, but every layer of governance that should have stopped it, in-house legal advice, the Cabinet approval process, the Ombudsman, the Administrative Appeals Tribunal, and Senate inquiries, was misled, bypassed, or rendered ineffective. A Federal Court declared the underlying decisions were never validly made, and a Royal Commission called the scheme neither fair nor legal. The result was not a technical failure but an institutional legitimacy collapse: billions in settlements, a Parliamentary apology, corruption findings against senior officials, and documented harm including suicides among vulnerable recipients.

Trajectory: Remediation has proceeded through forced court-ordered settlements (A$1.872B plus a pending A$475M Knox settlement) and 56 accepted Royal Commission recommendations, but recovery was externally compelled rather than self-initiated.

Inside the company · not a cell

Nothing quoted yet. A rate or a saving will sit here when it is on the record. It does not light the grid.

Initiative Map
CellInitiativeConfidenceEvidence
Q3AutomatorRoyal Commission, Volume 1 (7 July 2023)Income averaging applied automatically✅Page xxiv: averaged PAYG data was applied automatically where alternative information was not provided and accepted, with the aim of reducing and preferably removing human involvement. Page 131: ninety-five percent of staged reviews to 30 September 2016 were completed without any action by a compliance officer, and the default calculation was averaging.As of October 2026
Q3AutomatorPAYG Manual Compliance Intervention pilot (Jul 2015 – Jun 2016)Bridging program — partial automation of identification, manual verification still nominally present✅Pilot phase identifying ~860,000 cases of potential overpayment 2010–13 (Wikipedia, Sage analysis)As of April 2026
Q2TacticianPre-2015 manual compliance program (PAYG/data-match with case-worker verification)Routine debt-recovery process — Q2 case management with data-matching support✅~20,000 interventions per year; case workers manually verified information with recipient or employer before raising debt (Sage Open IS)As of April 2026
Q2TacticianReverse onus of proof on recipient"Customers had the opportunity to submit information about their income" — DSS/DHS framing to Ombudsman✅Recipients required to disprove a system-generated determination, sometimes for periods predating ATO retention requirements; "complex fact finding and data entry functions" shifted from department to the individualAs of April 2026
Q3AutomatorOnline Compliance Intervention (OCI) — debt calculation and notification (Jul 2016 – Feb 2017)"Automated data-matching scheme" / "compliance modernisation" / projected A$1.5 billion budget recovery✅"No meaningful human intervention in the calculation and notification of debts under the OCI phase" (Royal Commission Report at p. 477 via Pursuit); ATO PAYG annual income averaged across fortnights and treated as actual fortnightly earnings; ~20,000 debt notices/week capacity (~40x prior manual rate); ~783,000 interventions estimated for 2016–17 aloneAs of April 2026
Q3AutomatorSuccessor branding (Employment Income Confirmation Feb 2017 – Sept 2018; Check and Update Past Income Sept 2018 – May 2020)Continued compliance program — "helpful modifications" — more informative debt notices, improved interface, mechanisms to filter complex cases for manual processing✅Core mechanism of averaging-based debt calculation continued; modifications "did little to address the public outcry, because they did not rectify the system's core problems of speculative calculation of debts and reversed onus of proof"As of April 2026
Q2TacticianAAT review pathway (2016–22)"Robust review pathways exist" — implicit institutional claim✅431 AAT Tier 1 decisions questioned legal basis of averaging; some held averaging-based debts unlawful (notably 8 March 2017 finding that methodology was unlawful); decisions were not published; not appealed by DHS; not surfaced to SecretaryAs of April 2026
Q3AutomatorCommonwealth Ombudsman investigation (Apr 2017)"Independent investigation cleared the scheme" — used by ministers to defend Robodebt✅Found transparency/usability/fairness issues; did not address legality; later finding by Royal Commission that DHS and DSS officers "engaged in behaviour designed to mislead and impede the Ombudsman"; 2019 draft Implementation Report edited at departmental request to remove comments about legalityAs of April 2026
Q3AutomatorCabinet New Policy Proposal (2015)Routine compliance modernisation NPP estimated to recoup ~A$1.5 billion✅Did not disclose intended use of income averaging or the in-house legal advice that averaging was inconsistent with legislation; Royal Commission found NPP misleading; "as dishonest a document as the Commission has seen" used in relation to a related submissionAs of April 2026
Q3AutomatorRecovery operations"Budget repair" — A$1.5 billion projected savings✅Approximately A$1.73 billion in unlawful debts raised against ~433,000 people; A$751 million unlawfully recovered from ~381,000 people; ~648,000 group members in class actionAs of April 2026
Q3AutomatorDocumented harm to vulnerable populationsImplicit framing of recipients as "welfare cheats" / "drag on the national economy"🔶At least three known suicides connected to scheme; Royal Commission "confident these were not the only tragedies of the kind"; 663 vulnerable people (per Wikipedia summary of evidence) reportedly received debt notices and died soon after; debt notices issued to deceased persons and disability pensioners; sustained mental-health and stigma harm per RANZCPAs of April 2026
Q2TacticianPost-Robodebt remap (Nov 2019 onward)"Centrelink would no longer raise robo-debts calculated solely using income averaging without additional evidence"✅Federal Court consent orders in Amato; programmatic withdrawal May 2020; Prygodicz (No 2) settlement Jun 2021; Royal Commission Jul 2023; Government accepts 56 recommendations Nov 2023; APSC findings against 12 officials Sept 2024; NACC Operation Myrtleford Mar 2026As of April 2026
Published record

The lines this reading rests on. A date is the source date, when the page or the line names one.

The Royal Commission has found that “Robodebt was a crude and cruel mechanism, neither fair nor legal, and it made many people feel like criminals.

The Robodebt Scheme only came to an end in 2020 after the Federal Court found that it was unlawful in late 2019.

Commissioner Catherine Holmes AC SC presented the Report of the Royal Commission into the Robodebt Scheme to the Governor-General … on 7 July 2023.

Strength
The underlying data-matching mechanism worked exactly as designed, reliably flagging discrepancies between ATO and Centrelink records at massive scale.
Fragility
The system removed the human case officer from the decision loop entirely, so income averaging was treated as equivalent to evidence of actual fortnightly earnings with no case-level review.
Failure mode
Debts were raised and enforced against citizens using a calculation method that a Federal Court later ruled could never lawfully establish a debt, producing A$1.73 billion in unlawful debts and documented harm including suicides.
Governance ceiling
Pre-deployment legal warnings that averaging was inconsistent with legislation were never disclosed to Cabinet, and the Ombudsman, AAT, and Senate oversight gates were each misled or suppressed rather than allowed to function.
← Back to the Case Library