Deloitte Accused of Using AI in Preparing a Report for the Australian Government
Date: 2025-10-09 Author: Oliver Abernathy Categories: BUSINESS
Deloitte Australia found itself at the center of a scandal after it was revealed that generative AI tools were used in preparing a report for the Department of Employment and Workplace Relations (DEWR). The $440,000 document, published in July 2025, was found to be riddled with errors, inaccuracies, and references to non-existent sources.
An investigation revealed that a significant portion of the report was created by AI rather than by Deloitte specialists. Following the disclosure of these findings, Deloitte conducted an internal review and revised the document. The updated version of the report removed fictitious references, corrected footnotes, and updated the bibliography. Furthermore, the company officially acknowledged for the first time that it used generative models, including Azure OpenAI GPT-4o.
Despite the changes, the scandal sparked widespread discussion in Australian academic and professional circles. University of Sydney professor Christopher Raj noted that the errors in the document are typical of so-called "hallucinations" of artificial intelligence. He emphasized that the lack of transparency in the use of such tools calls into question the credibility of the recommendations outlined in the report.
DEWR representatives stated that the key findings of the report remained unchanged, but declined to specify whether Deloitte would be required to fully refund the funds or continue to cooperate with the agency. It is only known that the company agreed to pay the final contract fee, the amount of which was not disclosed.
In addition to the problems with the report itself, the audit uncovered internal weaknesses in DEWR's system. These included process errors, weak documentation, and overly strict compliance controls. These findings echoed an earlier report by the Commonwealth Ombudsman, which found that many cases of social security payment suspensions were unlawful.
Experts note that the incident has raised alarm bells regarding the reliability and ethical aspects of using artificial intelligence in government and corporate research. Deloitte, which actively uses AI in its work, now faces the risk of losing the trust of both clients and the public.
This incident marks one of the first times a major consulting firm has acknowledged that artificial intelligence was involved in the preparation of an official government document, intensifying the public debate about the limits of responsibility in the use of AI.
