PwC just got caught trying to pass AI slop as authentic research

[DragonSlayer101]

Facepalm: PricewaterhouseCoopers, one of the world's four largest accounting firms alongside EY, KPMG, and Deloitte, has landed in hot water after it was found to have published at least four reports riddled with AI hallucinations, fabricated citations, and fake footnotes. The reports were issued by PwC Middle East between 2024 and 2026 and were flagged as AI-generated by GPTZero, an AI detector and plagiarism checker.

The AI hallucinations were confirmed by multiple leading publications, including the Financial Times. The so-called "Thought Leadership" reports were intended to articulate PwC's insights into some of the fastest-growing businesses in the Middle East, including agentic AI, public services, and electric vehicles. However, the bizarre turn of events suggests that the reports themselves were AI-generated.

One of the contentious reports claimed that an internal PwC survey found that nearly 70% of CEOs in the region believed generative AI would redefine their business landscape in the near future. However, a footnote in the document linked to an article that did not cite any such survey.

Another footnote, this time in a cybersecurity report, included a link containing the tracking tag "utm_source=chatgpt.com," leaving very little doubt about its source.

In one particularly egregious case, a report cited a Medium blog post by a teenager with 280 followers as the sole source for a "real-world success story" about an AI deployment initiative by JPMorgan. According to the report, the initiative automated commercial loan agreement reviews, saving the company millions of dollars and thousands of hours of human work.

However, there was just one problem: JPMorgan's project actually took place in 2017, years before ChatGPT or any other public-facing AI chatbot was launched.

The string of errors, unfortunately, did not end there. One report on EVs cited a study on air quality in Riyadh that could not be traced to any journal, while another contained footnotes that led to broken webpages.

Another EV report claimed on at least three separate occasions that 90% of road accidents are caused by human error. However, the report cited two different studies for two of those claims and provided no attribution at all for the third.

Speaking to the FT, GPTZero policy analyst Paul Esau admitted that the data on road accidents isn't fake but argued that "no human is going to cite the same fact three times in two pages using three different sources."

Esau explained that "the chaotic signposting of source material is symptomatic of AI-generated research," adding that it is particularly ironic that PwC should make such blunders, given that the company markets itself as an authority on how to use AI responsibly and avoid exactly these types of errors.

It is worth noting that PwC rivals EY and KPMG were also caught passing off AI-generated reports as their own and had to retract those papers after they were flagged by GPTZero.

Like its rivals, PwC also acknowledged its mistake and vowed to change its policies to avoid similar faux pas in the future. In a statement, PwC Middle East said it "takes the accuracy of our published research seriously" and is "updating a limited number of supporting citations."

Permalink to story:

PwC just got caught trying to pass AI slop as authentic research

 

[pnntmp]

We know already that everyone uses AI.

What I don't see in the article is an estimation of the plausibility of the (obviously AI -assisted) research.
Where exactly is the research wrong (if it is), and why? This remains entirely unclear.

 

[Axeia]

PwC Middle East said it "takes the accuracy of our published research seriously"

Right...
We just posted AI generated garbage we didn't double check properly, but we take the accuracy seriously.

AI can be a useful tool, but generating whole reports without even checking if what was generated gets school students who cba doing their homework themselves caught. That's the level of 'serious' we're talking about.

The bar is a bit higher for an accounting firm where this likely was passed through the hands of multiple very well paid and educated people before it got published. If just copy pasting what AI generates was 'serious' enough there is some massive cost savings that can be made in the way of firing a whole bunch of people.

 

[Endymio]

Instead of blaming AI, why not blame the PwC employees who issued these reports without proper fact checking or research?

"PwC Middle East focuses its diversity, equity, and inclusion (DEI) policies on advancing gender parity, supporting localization goals, and ensuring fair talent pipelines. The firm actively integrates inclusive decision-making into its processes ...

...PwC's Inclusion First Framework fdocuses on racial and gender equity, cross-border inclusion, LGBT+ inclusion, and social inclusion....

• "... Features specific programs to elevate senior female professionals into leadership roles and improve emerging female talent trajectories...Integrates gender tracking across human resources processes (hiring, promotions, and pay) and supports regional frameworks designed to close the gender gap..."

... Localization requires shifting decision-making from HQ to local leadership so programs reflect the immediate cultural, social, and legal realities of the host country...."

 

[Gezzer]

Instead of blaming AI, why not blame the PwC employees who issued these reports without proper fact checking or research?

"PwC Middle East focuses its diversity, equity, and inclusion (DEI) policies on advancing gender parity, supporting localization goals, and ensuring fair talent pipelines. The firm actively integrates inclusive decision-making into its processes ...

...PwC's Inclusion First Framework fdocuses on racial and gender equity, cross-border inclusion, LGBT+ inclusion, and social inclusion....

• "... Features specific programs to elevate senior female professionals into leadership roles and improve emerging female talent trajectories...Integrates gender tracking across human resources processes (hiring, promotions, and pay) and supports regional frameworks designed to close the gender gap..."

... Localization requires shifting decision-making from HQ to local leadership so programs reflect the immediate cultural, social, and legal realities of the host country...."

Do they share the blame? Of course. But would the fact checking reduce the reliance on AI generated reports to the point where there was little to no savings to be had by using AI in the first place? I'd suggest most likely, and that's the problem in the end. AI is being heavily sold as a labour saving device, which means lower cost.

So companies are pushing their employees to utilize AI as efficiently and cheaply as they can. Especially as the real cost of AI is making itself felt. I can easily see companies that use AI to generate reports also using AI to fact check as well. After all the ones dictating the use of AI often have overconfidence on what AI can actually do due to all the surrounding hype.

The big problem as I see it is the myth that AI can do anything without error. When in fact it's a tool with limited real uses that requires human input or it will eventually go off the rails. And until upper management starts to realize this, stories like this will be more commonplace, not less. AI is not the universal answer, period...

 

[Endymio]

would the fact checking reduce the reliance on AI generated reports to the point where there was little to no savings to be had by using AI in the first place? I'd suggest most likely

For every report like this you read, there are ten thousand cases where AI produced output that was concise, accurate, and dramatically time-saving. As for comparing it to human output, one of my daughters is a senior executive for a Fortune 100 financial firm, and she regularly regales me with eye-popping tales of the absurd conclusions her analysts reach, sometimes with accounting errors off by multiple orders of magnitude.

And of course, the true elephant in the room is this: every year those AI models improve. Human analysts appear to going in the opposite direction.

 

[Hecate91]

Unsurprisingly they're taking the AI slop as real research, all of these financial firms are taking the AI slop seriously or else the AI bubble would've popped already.

 

[BobHome]

Unsurprisingly they're taking the AI slop as real research, all of these financial firms are taking the AI slop seriously or else the AI bubble would've popped already.

On the Plus side, it is creating more jobs for humans to sort through all the slop???

 

[Mr Majestyk]

We know already that everyone uses AI.

What I don't see in the article is an estimation of the plausibility of the (obviously AI -assisted) research.
Where exactly is the research wrong (if it is), and why? This remains entirely unclear.

I don't.