KPMG's October 2025 report on agentic AI in customer experience touted real-world wins at firms like UBS, but Financial Times checks with the named companies showed those details never happened. GPTZero's review of all 45 citations found only five fully intact, with most others paraphrased, invented, or unverifiable, prompting KPMG to remove the document while it investigates.
Amazing: KPMG wrote a report describing the successful use of AI by businesses. But the case studies turned out to be AI hallucinations. https://giftarticle.ft.com/giftarticle/actions/redeem/a2e02241-65c3-436e-9c47-fd0bd7981a9f
Nearly half the report's supporting claims rested on citations that turned out fake or garbled, a pattern GPTZero flagged as typical AI output rather than human research slips.
This case joins similar recent issues at other consultancies, leaving companies unsure how much to trust vendor-published AI adoption benchmarks without independent verification.
Many users condemned KPMG for publishing an AI adoption report with fabricated case studies, accusing the firm of deliberate lies, incompetence, and operating as a scam.
LOL
Amazing: KPMG wrote a report describing the successful use of AI by businesses. But the case studies turned out to be AI hallucinations. https://giftarticle.ft.com/giftarticle/actions/redeem/a2e02241-65c3-436e-9c47-fd0bd7981a9f
yikes
Amazing: KPMG wrote a report describing the successful use of AI by businesses. But the case studies turned out to be AI hallucinations. https://giftarticle.ft.com/giftarticle/actions/redeem/a2e02241-65c3-436e-9c47-fd0bd7981a9f
lol
FT Exclusive: A KPMG report on how AI is being used by businesses across the world exaggerated adoption of the technology with bogus case studies that appear to have been based on AI hallucinations. https://ft.trib.al/z44Q3aR
@anneapplebaum GROK is becoming more and more "vague," and non-responsive.
GROK tries to 'misdirect,' almost everytime I prompt a question. Then "ERROR"
It claims it has 'no information,' on wide areas of world events, individuals, and "Epstein" references.
@anneapplebaum Wow, I can’t believe nobody inside KPMG thought to fact check it.
@Aligned_SI @anneapplebaum No, the lesson is it cannot be trusted, at least right now. And relying on it is dangerous and expensive.
And to ignore that will cause very expensive deleterious consequences.
Cc: @MarkJCarney @gary_srp @SylviaJonesMPP #AI
@anneapplebaum It's garbage. It's anti-civilization. It's going to destroy us if we don't stop it.
@anneapplebaum
@AndrewHazlett @anneapplebaum Yes the AI ouroboros consuming itself! And @KPMG, where is the human oversight? Judgement? What happened to the human-in-the-lead strategy? All of this is critical for maintaining trust.
@anneapplebaum Sounds like some severe mental illness. If you can correct for this, one can have a good experience using AI effectively.
@anneapplebaum KPMG is a criminal organization. But so is all of Big 4 accounting firms.
@Noahpinion 😂
@anneapplebaum Never underestimate the bravado of dumb people with powerful tools.
@anneapplebaum Write about kpmg, Deloitte, McKinsey, consultants, etc are plundering government funds
@anneapplebaum You would think they at least would hire a good prompt engineer
@Noahpinion Why are we worried about jobs being lost if KPMG still manages to bill customers millions while doing this
@anneapplebaum You know I can’t see this working out well for humans or humanity.
@Aligned_SI @anneapplebaum "The dishwasher I sell sometimes doesn't clean dishes. The lesson isn't that the dishwasher can't be trusted to clean dishes, it's that organizations need processes to ensure humans know when to eat off of the dishes in the dishwasher"
@anneapplebaum
@anneapplebaum You can’t make this sh*t up.
