Deloitte Refunds Australia After AI-Generated Report Errors

Thumbnail Image

The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.

Deloitte issued a partial refund to the Australian government after its AI-assisted report on the welfare compliance system contained fabricated citations and errors due to hallucinations from OpenAI's GPT-4o. The incident raised concerns about AI reliability in official documents, though Deloitte claimed the main findings remained unchanged.[AI generated]

Why's our monitor labelling this an incident or hazard?

The event explicitly involves an AI system (generative AI language model) used in the creation of a report that contained fabricated and false information (hallucinations). This misinformation in an official government report can be considered a violation of legal obligations and harms the integrity of governmental decision-making, which fits the definition of harm to rights and obligations under applicable law. The AI system's use directly led to these harms, making this an AI Incident. The partial refund and report revision are responses but do not negate the incident classification.[AI generated]
AI principles
Robustness & digital securitySafetyTransparency & explainabilityAccountability

Industries
Government, security, and defence

Affected stakeholders
GovernmentBusiness

Harm types
Economic/PropertyReputational

Severity
AI incident

Business function:
Compliance and justice

AI system task:
Content generation

In other databases

Articles about this incident or hazard