A new report that was supposed to be a shining panegyric to how useful AI is was caught loaded with fake claims that appear to be AI hallucinations, the Financial Times reports — a blunder that may have unintentionally demonstrated AI’s most compelling use case: bullsh*tting your job.
Titled “Redefining excellence in the age of agentic AI,” the report was released October by KPMG, one of the big four consulting firms, making the mishap an all the more embarrassing indictment of elite Wall Street professionals’ mindless enthusiasm for the tech.
It claimed that numerous big organizations are already making extraordinary use of the tech, with the takeaway that you should be deploying AI, too, lest you become one of those dinosaurs that still relies on the fleshy circuitry of the brain to synthesize information.
The global wealth manager UBS “integrates AI agents across investment advisory, risk management and compliance monitoring,” it stated. The Swiss Federal Railways has AI agents that “help users plan, book, and optimize journeys based on preferences, real-time conditions and carbon impact, turning SBB into a holistic mobility orchestrator,” it further asserted. And the Transport for London was using AI agents “to predict and manage congestion, personalize commuter updates and co-ordinate multimodal transport.”
AI evangelists, never above shameless boosterism and misleading claims, would desperately hope for any of these things to be true. But as the FT was tipped off by AI detecting startup GPTZero, they’re not. A UBS spokesperson told the FT that the claims about AI agents were “factually incorrect.” A Swiss railway spokesperson echoed that the claims made about its AI usage were “not accurate.” And a spokesperson for the London transit system called the report’s assertion “misleading.”
The provenance of another false claim, that the NHS Greater Manchester uses AI agents to do everything from triage patients and predict hospital readmissions, is a classic game of AI telephone. It appeared to be based on a press release, with footnotes citing a communiqué about an AI tool designed to combat lung cancer — which had nothing to do about an AI agent performing the above hospital tasks, according to the FT.
KPMG pulled the report from its websites after it was alerted about the bogus claims, but the damage is done: its findings were already cited by industry publications and a major Czech newspaper, GPTZero found.
For AI fabrications to be trickling down from a consultancy of KPMG’s stature “poison the well of information,” GPTZero CEO Edward Tian told the FT, increasing “the risk of second-hand hallucinations.”
Like in many white collar fields, AI has caused a bit of an existential panic at consulting firms, which fear that it could outmode their profession. But, free thinkers all, their response generally has been to embrace it like everyone else is. McKinsey reportedly deployed 12,000 AI agents across the company last year, shortly after laying off some five thousand employees.
This is far from the only example of how these efforts can backfire. Lawyers, including those at white shoe law firms, have been repeatedly caught submitting court filings riddled with fake case law and misattributed quotes that turned out to be hallucinated by an AI tool, incurring the admonishment of the judge.
More on AI: Furious Judge Cancels Entire Trial After Finding Out Lawyers on Both Sides Used AI
The post Consulting Firm’s Report on How Awesome AI Is Found to Contain Idiotic AI Hallucinations appeared first on Futurism.
