AI used to cheat on an AI Test - 28 KPMG employees caught, along with a partner who is fined $10k

KPMG Australia has been dealing with 28 cases of staff using artificial intelligence to cheat on internal exams since July, including a senior partner who faces a financial penalty of more than $10,000 for breaching internal AI-use policies during mandatory training.

These incidents come to light as the firm positions itself as a leader in AI consulting, and after years of prior cheating concerns in internal tests between 2016 and 2020, which prompted the business to overhaul how it designs, monitors, and polices staff assessments.

• https://theaussiecorporate.com/blogs/pickandscrollnews/kpmg-faces-ai-cheating-scandal

NOT THE FIRST TIME

KPMG has a well-established pattern of exam-related integrity failures, not only in the recent AI misuse scandal, but also historically:

• Major exam‑cheating scandals in 2019 (US),
• Massive five-year cheating operation uncovered in the Netherlands (2017–2022),
• Earlier exam‑integrity problems recorded in Australia (2016–2020),
• Ongoing issues culminating in AI exam cheating in 2026.
• https://www.sec.gov/newsroom/press-releases/2019-95
• https://tax.thomsonreuters.com/news/us-audit-regulator-imposes-record-penalty-on-kpmg-netherlands-for-exam-cheating-by-hundreds-of-professionals/

A Modern Irony in Professional Services

• In a development that has captivated the global business and technology community, a senior partner at KPMG Australia has been fined A$10,000 for using generative artificial intelligence tools to cheat on a mandatory internal training exam—an exam specifically designed to assess ethical and responsible use of AI.
• The story, first reported on 16 February 2026, highlights the growing tension between emerging technologies and corporate governance.
• The exam required the completion of an AI ethics and responsible‑use training module, part of the firm’s broader initiative to standardise understanding of AI in audits, advisory work, and data‑driven decision‑making. However, this mandatory step toward trust and competency backfired when the partner uploaded the course material into an external AI platform to generate correct answers—violating explicit firm policy.

How the Incident Unfolded

• The misconduct occurred during a July 2025 AI training assessment.
• The course materials explicitly encouraged staff to download a reference manual but strictly prohibited inputting those materials into external AI systems. Internal monitoring technology—implemented by KPMG in 2024 to detect improper AI use—flagged the suspicious activity in August 2025.

An internal investigation concluded that the partner had breached firm policy. As a result:

• The partner was fined A$10,000 in future income,
• Required to retake the training exam, and
• Obliged to self-report the breach to the professional body, Chartered Accountants Australia and New Zealand, which has opened its own investigation.
• This case is not isolated. KPMG confirmed that 28 staff members have been caught using AI tools to cheat in internal assessments during the current financial year. Most were junior or mid-level employees, but the partner’s seniority drastically increased regulatory scrutiny.  

Regulatory and Political Scrutiny Intensifies

• The incident came to light during an Australian Senate inquiry into corporate governance. Greens Senator Barbara Pocock criticised the “extremely disappointing” behaviour and pointed out weaknesses in existing oversight mechanisms, calling the system “toothless” for allowing such breaches to go largely undeclared.  
• Meanwhile, the Australian Securities and Investments Commission (ASIC) was notified but indicated that further action would depend on the outcome of professional disciplinary proceedings.
• This is notable given that audit partners—particularly registered company auditors—carry enhanced responsibilities due to their access to sensitive financial information and their role in upholding trust in the audit process.

KPMG’s Response: A Struggle to Contain AI Misuse

• KPMG Australia CEO Andrew Yates acknowledged the challenges of regulating generative AI in a rapidly evolving technological environment.
• According to Yates, the firm has been “grappling with the role and use of AI” due to society’s rapid adoption and the increasing ease with which powerful AI tools can be misused.  

To combat the issue, KPMG has:

• Implemented AI-monitoring technologies,
• Launched a firm-wide education campaign,
• Introduced stricter controls to block AI access during assessments, and
• Committed to publicly reporting AI-related misconduct cases in its annual results.

These system enhancements were introduced after widespread cheating in internal training modules was first detected between 2016 and 2020.  

A Broader Trend: AI Cheating on the Rise

• The event comes amid wider concerns across professional industries about the increasing sophistication of AI-based cheating tools. For example, the Association of Chartered Certified Accountants (ACCA) cancelled remote exams in 2025, citing inadequate safeguards against AI-driven cheating.  
• Other firms, including Deloitte, have similarly faced embarrassment after AI-related errors appeared in official government reports.  
• This reflects a growing transparency crisis: as organisations demand more efficiency through AI, they must simultaneously enforce stricter controls to ensure employees actually understand the technology rather than outsourcing their own training to it.

Why This Case Matters

This incident is about far more than a single partner's lapse in judgment. It raises critical questions about:

• Professional integrity in an AI-driven world
• Regulatory preparedness for AI misuse
• The reliability of AI competency certifications
• Internal governance in firms that heavily rely on AI-supported work

KPMG’s situation illustrates the immense pressure on firms to adapt quickly while maintaining ethical standards—especially when the tools used to enhance productivity can just as easily undermine it.

And for a Big Four firm advocating AI-powered efficiencies to clients, the breach is doubly damaging: it reveals internal weaknesses. It opens the door to external criticism about whether staff truly possess the technical proficiency the organisation claims to have.

Conclusion

• The A$10,000 fine imposed on the KPMG partner underscores a paradox of the digital age: AI is both a powerful tool and a growing risk. As more organisations integrate AI into everyday processes, incidents like this are likely to continue unless oversight, training integrity, and ethical norms keep pace with the technology itself.
• This case is a wake-up call—not just for KPMG, but for industries worldwide navigating the rapidly evolving AI landscape.

History of cheating!!!

1. United States (Major scandal, 2019)

KPMG US was fined $50 million by the SEC in 2019 after regulators found:

• Widespread cheating on internal training exams, including sharing answers and manipulating scores.
• Cheating involved numerous audit professionals.
• The misconduct was serious enough that KPMG was required to hire an independent consultant to overhaul ethics and integrity controls. [sec.gov]

This was one of the most severe ethics failures ever recorded for a Big Four firm.

2. KPMG Netherlands (Record-breaking scandal, 2017–2022)

In April 2024, the PCAOB fined KPMG Netherlands $25 million, the largest exam-related penalty in PCAOB history, after finding:

• Widespread cheating by more than 500 professionals on mandatory internal training exams between 2017 and 2022.
• Cheating reached senior leadership, including the firm’s Head of Assurance.
• Some leaders misled regulators and failed to disclose known misconduct.
• The firm failed to maintain adequate quality control and an ethical culture. [tax.thomso...euters.com]

This scandal highlighted systemic failures across multiple years and levels of seniority.

Overall Conclusion

KPMG has a well-established pattern of exam-related integrity failures, not only in the recent AI misuse scandal, but also historically:

• Major exam‑cheating scandals in 2019 (US),
• Massive five-year cheating operation uncovered in the Netherlands (2017–2022),
• Earlier exam‑integrity problems recorded in Australia (2016–2020),
• Ongoing issues culminating in AI exam cheating in 2026.

This long-running pattern suggests systemic weaknesses in ethics culture, test administration, and internal controls across several KPMG member firms.

Source

• [hcamag.com] - https://www.hcamag.com/au/specialisation/employment-law/kpmg-partner-fined-over-ai-exam-breach/565406
• [cnbctv18.com] - https://www.cnbctv18.com/technology/kpmg-australia-partner-fined-for-using-ai-to-cheat-in-ai-ethics-test-ws-l-19850963.htm
• [financialexpress.com] - https://www.financialexpress.com/life/technology-kpmg-australia-partner-fined-for-cheating-ai-ethics-training-exam-using-ai-4144532/
• [thestatesman.com],  - https://www.thestatesman.com/technology/the-ultimate-irony-kpmg-partner-fined-10k-australian-dollars-for-using-ai-to-clear-ai-test-1503557469.html
• https://www.sec.gov/newsroom/press-releases/2019-95
• https://tax.thomsonreuters.com/news/us-audit-regulator-imposes-record-penalty-on-kpmg-netherlands-for-exam-cheating-by-hundreds-of-professionals/
• Australian Financial Review https://www.afr.com/companies/professional-services/kpmg-partner-fined-for-using-ai-to-cheat-on-an-ai-test-20260215-p5o2f2
• Going Concern https://www.afr.com/companies/professional-services/kpmg-partner-fined-for-using-ai-to-cheat-on-an-ai-test-20260215-p5o2f2
• CFO Dive https://www.cfodive.com/news/deloitte-ai-debacle-seen-wake-up-call-corporate-finance/802674/
• KPMG https://assets.kpmg.com/content/dam/kpmgsites/uk/pdf/2026/01/reimagining-controls-assurance-with-ai.pdf.coredownload.inline.pdf