Artificial intelligence is reshaping the landscape of fraudulent expense reporting, as new image-generating AI models are making it easier for employees to submit fake receipts. According to software provider AppZen, approximately 14% of fraud attempts now involve AI-generated receipts, a significant increase from previous years.
The emergence of advanced image-generation models from prominent AI groups like OpenAI and Google has led to a surge in AI-generated receipts being circulated within organizations. Notably, fintech company Ramp detected over $1 million in fraudulent invoices within just 90 days using its new software.
A survey conducted by expense management platform Medius revealed that about 30% of financial professionals in the US and UK noticed a rise in falsified receipts following the introduction of OpenAI’s GPT-4 model last year.
Chris Juneau, Senior Vice President and Head of Product Marketing at SAP Concur, a leading expense platform processing millions of compliance checks monthly with AI, emphasized the increasing sophistication of these AI-generated receipts, cautioning against relying solely on visual inspection.
Previously, creating fake documents required photo editing skills or engaging online services. However, the availability of free and user-friendly image generation tools has enabled employees to fabricate receipts swiftly by providing simple text instructions to chatbots.
Source: Ars Technica