Artificial Intelligence Against Artificial Intelligence: Layoffs and the Fraud War in Banking

Standard Chartered Bankası CEO Bill Winters' description of employees to be replaced by artificial intelligence as "lower-value human capital" drew massive backlash. This statement emerged as part of a plan to lay off fifteen percent of office staff to meet the bank's profit targets. After receiving intense criticism from many, including the former Singapur head of state, Winters attempted to retract his remarks. However, this incident also laid bare how rapidly and ruthlessly the use of AI is spreading in the financial sector. The banker simply voiced aloud the disturbing truth that everyone knows but no one says: For years, compliance officers had already been expected to work like inefficient computers.

The job of bank employees tasked with combating financial crime is not as exciting as that of detectives in movies. On the contrary, thousands of people sit in small cubicles in Varşova or Bengaluru, having to check transactions flagged as suspicious by automated systems. These employees' primary duty is to confirm that ninety-nine percent of the flagged transactions are actually normal and escalate the rest to upper management. Upper management, in turn, decides that almost all of these escalated transactions are non-problematic, terminating the process. This is an extremely exhausting, repetitive, and costly process, yet it appears to be surprisingly ineffective at stopping crime.

The crime economy continues to grow despite all the penalties imposed and laws enacted over the years. The fundamental goal of banks' efforts to combat financial crime is not actually to prevent crime, but to avoid massive fines. Standard Chartered had also been hit with hefty fines in both Amerika Birleşik Devletleri and Birleşik Krallık in the past due to inadequate controls. For this reason, banks are eagerly embracing this technology, as long as they are certain that AI can check the boxes in a way that satisfies regulators. However, this box-checking approach, focused on fulfilling regulatory rules, is far from providing real security.

The worrying aspect of the matter is that AI has already become incredibly adept at perpetrating fraud. Crime syndicates, just like banks, want to reduce back-office expenses and are using AI to automate small-scale operations. Blockchain analytics firms like TRM Labs point out that AI is being used in phishing and scams known as "pig butchering," which target vulnerable individuals by mining social media data. Because AI-powered bots can maintain persuasive dialogues without getting tired or making mistakes, fraud has become much easier to scale. Last year, an enormous increase of approximately five hundred percent was recorded in AI-powered fraud activities within the cryptocurrency world.

In light of all these developments, the replacement of bank employees with AI continues to be criticized. The main problem is that instead of laying off these personnel, banks do not train and use them as a line of defense against AI-powered crimes. Warren Davidson, the head of the illegal finance subcommittee in the Temsilciler Meclisi, praised the attacks on corporate transparency legislation, completely misunderstanding the situation. It is clear that an AI model not trained with reliable data can be no more beneficial than the existing bureaucratic system. It is impossible for a computer that does not know who owns what to detect financial crimes, and therefore, the issue focuses more on the quality of the data than on the technology itself.