AI: The next big scandal

Asked whether the introduction of AI tools would make them more or less likely to trust the result of an external financial audit, 87% of audit committee members said they would trust AI over people. That’s not surprising: The last decade has seen more than its fair share of audit-related scandals, most of which have been caused by very human factors, like greed or criminal intent. But before the leaders of professional services firms rush to announce yet more cuts in their graduate intake, we might do well to examine the risks AI poses to trust, too.  

A short history of professional scandals 

Scandals are nothing new.  

My research into this took me all the way back to the 1856 case of the Great Northern Railway in the UK, whose auditor failed to notice that one of the railway’s employees had created and sold more than £220,000 of fictious stock. Google Gemini estimates that that sum would be worth around £7.5m today. The first society of accountants was founded in 1854, so the auditor’s negligence occurred just two years after the beginning of the audit industry as we know it today. 

Just over 20 years ago, another audit scandal, at the energy trading company, Enron, engulfed Arthur Andersen, whose audit team had signed off on various misleading statements and then tried to destroy the evidence. The case was unusual because it—obviously—wasn’t the first scandal to hit the audit industry or indeed Arthur Andersen, but it did destroy the firm when previous scandals hadn’t. Researching the case at the time, I concluded that two factors tipped what could have been a run-of-the-mill affair into something far more serious. The first was that, with all the publicity surrounding Enron’s collapse, clients and the regulator started to join the dots: Actions that might have been attributed to “rogue” partners started to be seen as something more systemic. Second, two up-and-coming politicians, Senator Paul Sarbanes and U.S. Representative Michael G. Oxley, were willing to pick a fight with a major firm. (For younger readers: They won.)  

Although the causes of scandals in the professional services industry have varied, their impact has therefore been determined by other factors. Firms can contain the potential fallout of a scandal by demonstrating that it only involved a small number of people, acting independently. Things only get serious when people start to see deliberate and systemic conspiracy. Even then, if firms can avoid significant and relentless political scrutiny, the lack of media coverage will deprive a scandal of the oxygen it needs to thrive. Everyone moves on.  

AI-related scandals may be different 

It seems to me that AI-related scandals won’t be caused by criminal conspiracy, but by incompetence. Pouring over the entrails of future scandals, we won’t find secret emails encouraging sinister use of ChatGPT or cabals of partners in far-flung parts of the world advocating deliberate misuse of data. Much more likely, it will have been human error—the result of someone working late, cutting corners, or trying to meet impossibly tight deadlines. The latter can still be seen as systemic: “I’m not surprised to hear that Firm X has been misusing AI,” we’ll hear future clients say. “I always thought they were a bit slipshod and the partner was never around to review things properly.”  

The difference between AI- and non-AI-related scandals is that, precisely because there won’t be a conspiracy, containment will be much harder. Arthur Andersen collapsed because the actions of a small group of people were perceived to be the tip of a cultural iceberg. But in AI-related scandals, there will be no “partner zero”, the evil genius whose ambition leads others astray. AI-related scandals will emerge in many parts of a firm at once, driven by corporate cultures that don’t give people time to think or that have only a casual relationship with facts. When the culture is to blame, it’s hard to find people to blame. 

As for scrutiny, it will be hard to avoid. Since the arrival on the scene three years ago, generative AI has captured the imagination of a generation, triggered a level of investment not seen since the early days of the internet, and promised business leaders and politicians alike that it’s the solution to almost every insoluble issue in today’s world. The laws of hubris would tell you that this was going to come crashing down at some point, and there’s a certain amount of schadenfreude in watching overblown egos being punctured. Most non-AI-related scandals never develop to a point where politicians and the media took them seriously. Theoretically serious, they simply never get noticed. That’s much less likely to happen here. 

So what? 

I’d argue that all this raises some important questions for professional services firms. It’s pretty much inevitable that there will be many more instances of cavalier AI usage, most of which will be picked up by journalists. Firms will struggle to demonstrate that their quality control procedures can cope. Clients might increasingly forbid the use of AI in any capacity, destroying the value firms hope to earn from the investment they’re currently making in this technology. Worst of all, the professional services firms’ reputations for delivering high-quality work, produced by experts, could evaporate.  

There’s been a lot of discussion about whether AI will destroy the professional services industry. It could—but perhaps not in the way we expected. 

This is our final Leadership Market Update of 2025. Thanks for reading! We’ll be back in 2026 with more insights. If that’s too long to wait, you can browse the archive of past articles or check out our podcasts. 

What can firms do next? 

Source helps leading professional services firms understand how the growing influence of AI and other important trends are reshaping the industry. To find out more about our expert-led briefings on the latest developments, get in touch.