AI Is Moving Money: Banks Rethink Security and Liability

In the United States, losses reported from AI-related cybercrime in 2025 topped $893 million, across 22,364 complaints. As scammers use AI to impersonate banks and fake identities, and as AI tools begin to check accounts and handle transactions for customers, banks are being pushed to rethink their security systems.

Cheaper, Faster Fraud

The FBI's Internet Crime Complaint Center (IC3) separately tallied complaints that involved AI in its 2025 annual report. The reported losses came to $893,346,472. That figure covers only cases where AI was reported as part of the crime, so it does not capture every AI-related offense in the US. How much AI contributed to each case, and the actual scale of the damage, also varies from one case to the next.

AI-powered cybercrime cuts the time and cost scammers need to deceive victims. Criminals can generate voices that sound like real people and produce large volumes of messages that pose as bank staff. They can also automate the work of gathering information on a specific target and building a conversation that fits the situation.

Alongside this, technology that lets AI act for consumers, checking accounts and processing financial transactions, is advancing. That raises the stakes for how banks protect accounts. Control is expanding to cover transaction approval, execution records and even emergency shutdown. South Korea has also begun enforcing financial-sector AI guidelines and is putting in place a framework for proxy transactions, where an agent acts on a customer's behalf.

Why It Matters

As AI starts to move money on customers' behalf, banks must decide who approves each transaction, how actions are recorded and who bears responsibility when something goes wrong.

Read the original article in Korean: K trendy NEWS

댓글

이 블로그의 인기 게시물

Incheon Explores Cleaner Tap Water and Stream Restoration

체이스 홀이 커피로 그린 흑인 광부들, 우리 시대 미술이 되살린 노동의 역사