Skip to content

Robodebt

An Australian welfare debt scheme that relied on averaging annual income across fortnightly benefit periods. Income averaging stopped in November 2019; refunds, litigation and a royal commission followed.

Case studySources & confidence: Qualified confidence

Checked 2 October 2026 · AI-assisted review

Checked chronology, settlement categories and the nature of the 2025 legal announcement against Australian government sources.

What this check doesn’t establish

  • No total settlement figure is given because debt cancellation, refunds and compensation are different categories.
  • Individual harms, causal attribution and the full Royal Commission evidence were not independently reviewed.

Confidence applies to the scope above, not every claim made by the underlying source. How confidence works

The scheme used annual tax income to infer income over benefit-reporting fortnights, producing debts for recipients to contest. Services Australia identifies the relevant income-averaging period as July 2015 to November 2019.

The original class action settlement was approved in June 2021. Refunds and settlement payments were separate, so describing the combined debt relief as a single cash settlement is misleading. The September 2025 announcement concerned a class action appeal settlement, not a second class action. The Royal Commission reported in July 2023.

Record history

Catalogue source: knowledge/cases/au-robodebt.md

Generated by claude-code/claude-opus-5 on 2026-07-26

esc
Try
Cybersyn
why law should be legible
AI agents playing Civilization
Jump to
WritingInteractivesLibrary catalogueLabWorkAboutContact
Settings
Switch to dark theme