Artificial intelligence (AI) is reportedly helping employees scam their bosses with phony expenses receipts.<\/p>\n
The rise of new image generation models from companies like Google and OpenAI have led to a wave of AI-generated receipts submitted in-house, the Financial Times (FT) reported<\/a> Sunday (Oct. 26), citing interviews from leading expense software platforms.<\/p>\n For example, software provider AppZen<\/a> said fake AI receipts accounted for roughly 14% of all fraudulent documents submitted last month, compared to zero last year. The FinTech Ramp<\/a> said its new software caught more than $1 million in fraudulent invoices within 90 days.<\/p>\n And around 30% of the British and American financial professionals surveyed by expense management platform Medius<\/a> reported they had seen an uptick in falsified receipts after the launch of OpenAI\u2019s GPT-4o in 2024.<\/p>\n \u201cThese receipts have become so good, we tell our customers, \u2018Do not trust your eyes,\u2019\u201d said Chris Juneau<\/a>, senior vice-president and head of product marketing for expense platform SAP Concur<\/a>, which processes more than 80 million compliance checks per month using AI.<\/p>\n Several platforms pointed to a significant jump in the number of AI-generated receipts following the debut of GPT-4o\u2019s improved image generation model<\/a> in March. OpenAI told the FT that it takes action when its policies are violated and its images contain metadata that indicate they were created using ChatGPT.<\/p>\n Advertisement: Scroll to Continue<\/p>\n The FT noted that there was a time when someone would need photo editing skills \u2014 or the willingness to pay an online vendor \u2014 to create a fraudulent document. Now, free and accessible image generation software allows employees to quickly falsify receipts in seconds by giving simple text commands to chatbots.<\/p>\n Expense management platforms showed the news outlet several receipts featuring realistic details like wrinkles in paper, detailed itemization in line with real-life menus, and signatures. It\u2019s part of a larger problem, as PYMNTS wrote earlier this year.<\/p>\n \u201cGenerative AI has transformed financial fraud<\/a>, making it more sophisticated and harder to detect,\u201d that report said. \u201cWhile businesses harness its capabilities for efficiency, cybercriminals exploit it to craft convincing fraudulent schemes.\u201d<\/p>\n Modern AI-driven attacks aren\u2019t confined to phishing emails, as fraudsters now use voice cloning and deepfake<\/a> videos to impersonate executives, duping employees into authorizing large fund transfers.\u00a0<\/p>\n \u201cThe accessibility of generative AI tools means that even low-level scammers can produce high-quality forgeries, blurring the lines between genuine and fake communications,\u201d PYMNTS added.<\/p>\n