Deloitte has been in the news recently for something many find hard to believe a firm considered one of the Big Four accounting firms in the world could possibly commit. 

What was Deloitte commissioned to do? 

Well, in simple terms, Deloitte was hired by the Albanese government’s Department of Employment and Workplace Relations (DEWR) to check how the government’s welfare penalty system works — both the rules that decide when jobseekers should be penalised and the computer system that automatically applies those penalties — and then suggest ways to make the whole process fairer and more effective. 

Instead of conducting a 'thorough research study' using the whopping $440,000 the Albanese government paid, the big consulting firm used AI to help develop parts of the report. The report was later found to contain fake references and footnotes, including a fabricated quote from a federal court judge.

The incident has sparked widespread opinion, with those who have long shown their cynicism toward AI using the episode as a yardstick to insist that the technology is no good and cannot be trusted. 

But their discredit of the new technology, judging by history, is misguided — and, in fact, myopic.

In 1975, Kodak was actually the first company to invent the digital camera. However, fearing that it would disrupt their film business, they dismissed it. The outcome, as we know today, is that competitors like Canon and Sony embraced digital technology — and Kodak collapsed. 

We also know the story of how, in 2000, Netflix’s founders offered to sell the company to Blockbuster for $50 million. Underestimating the power of streaming technology and sticking to physical rentals, Blockbuster — which made a whopping $800 million that year — laughed them off. 

The rest, as they say, is history. 

My point is this: shaming Deloitte for using AI to aid in authoring that report is plainly misguided.

If Deloitte does not adopt AI, the simple truth is that others — like EY (Ernst & Young), KPMG, and PwC (PricewaterhouseCoopers) — will. And in no distant time, Deloitte would be out of operation. 

Every major technology has suffered from early misuse — not through malice, but misunderstanding. Kodak with digital cameras. 

Blockbuster with streaming. Newspapers with the internet. Deloitte’s AI mistake belongs to that lineage — not proof that AI is flawed, but evidence that we’re still learning how to wield it correctly. 

I have long advocated the use of AI as a co-creator. And what this means is that when creating with AI, you can’t afford not to fact-check its responses — just as an experienced editor cannot afford to publish a piece of writing without verifying its factual accuracy. 

This is where Deloitte failed. 

Even Dr. Christopher Rudge of the University of Sydney, who first flagged the report as containing “hallucinations,” refused to discredit the report’s conclusions or regard it as illegitimate.