PwC published "thought leadership" reports on AI and electric vehicles riddled with fake footnotes, misattributed claims and unverifiable information, according to an investigation by GPT Zero researchers verified by the Financial Times. The AI hallucinations appeared in four reports produced by PwC Middle East over the past two years, designed to drum up consulting work in the region.

Many citations link to web pages that don't contain the purported evidence or don't exist. An academic paper on air quality in Riyadh appears entirely hallucinated, with no trace in the referenced journal. A teenage blogger with 280 followers on Medium is cited as PwC's source for a JPMorgan "real world success story" of agentic AI from 2017, years before generative AI existed.

PwC Middle East said it "takes the accuracy of our published research seriously and is updating a limited number of supporting citations." Researcher Paul Esau noted tell-tale signs of AI generation, including a claim about human error causing 90% of traffic accidents cited three times in two pages with three different sources. One footnote URL even included "utm_source=chatgpt.com."

The findings risk embarrassing consulting firms marketing themselves as advisers on responsible AI adoption. GPT Zero's earlier investigations led rivals EY and KPMG to retract reports with similar hallucinations, highlighting a broader pattern across the Big Four.