A group of academics has issued an apology after unwittingly using Google Bard to generate erroneous claims about the big four consultancy firms in a formal submission to the Australian Parliament. This incident emerged during a parliamentary review of the professional conduct within the consultancy industry, which could potentially lead to the restructuring of the big four: Deloitte, KPMG, PwC, and EY.
The academic group submitted what they asserted were case studies revealing improper auditing practices by these firms. However, these case studies were fabricated by Google Bard, containing several serious inaccuracies. For instance, one claim falsely implicated KPMG in the wage theft scandal involving 7-Eleven, while another alleged that Deloitte faced a lawsuit from a construction company for faulty auditing practices. Additional mistakes included allegations against Deloitte for falsifying accounts related to the café chain Patisserie Valerie.
Following objections from the big four regarding these allegations, the scholars expressed regret and attributed the mishap to the use of artificial intelligence. Emeritus Professor James Guthrie articulated their remorse in a letter to the Australian Senate, stating, “As accounting academics, we have a deep commitment to serving the public interest and ensuring accountability and transparency in our sector. For this reason, we wholeheartedly apologize and express our profound regret. I have come to realize that AI can produce outputs that, while sounding authoritative, may be incorrect, incomplete, or biased.”
Moreover, Senator Richard Colbeck, who is overseeing a separate inquiry into the consulting industry, emphasized the gravity of the situation. He remarked that this incident highlights the potential chaos that can ensue when fundamental facts are blatantly incorrect.
This episode echoes a recent case in New York where a lawyer utilized ChatGPT to research fictitious legal cases, which he subsequently submitted to court. As a result of this misdeed, he and his partner faced a $5,000 fine, prompting an official apology.
The proliferation of artificial intelligence in various professional domains raises important questions about the reliability and accountability of information generated by such technologies. As AI continues to evolve and integrate into different fields, it is crucial for professionals to approach its outputs with discernment and diligence. Building awareness around these issues will be essential as industries adapt to the realities of AI and its impact on the integrity of their work.