Accenture accused of publishing major research full of AI hallucinations — paper featured as many as 15 fictitious citations and case studies

Accenture, a major consultancy firm, is facing criticism for publishing a research paper that contains numerous AI-generated inaccuracies. The paper, titled "Leapfrogging Development: Unlocking Africa’s Potential in Global Value Chains with Artificial Intelligence," was co-authored with IE University in Madrid. The paper has been identified by AI-detection software GPTZero to have 15 AI hallucination inaccuracies, along with four flawed citations.

GPTZero, which has previously flagged similar issues in reports from other large organizations, claims that AI hallucinations pose a significant danger to researchers, academics, consultants, students, and anyone who relies on the information found online.

The Accenture-linked research paper is not the only concern raised by GPTZero. The company was found to have produced three documents containing a combined total of 31 AI-generated hallucinations, along with numerous fabricated claims. GPTZero's automated pipeline and Hallucination Detector were used to uncover these issues, with a real human verifying each hallucination to ensure accuracy.

When questioned by The Telegraph, an Accenture spokesperson attributed the problems to human error rather than AI-generated hallucinations. The company has stated that they will correct the issues before releasing a revised version of the paper. Despite being a collaborative effort, the research paper was produced beyond Accenture's standard in-house process, highlighting ongoing challenges the firm faces when it comes to incorporating AI into its work.