As digital banking gains ground, banks are confronting exotic new fraud threats including deepfakes and synthetic identities. And while online banking brings unprecedented forms of value to business payments, it is also opening new avenues for B2B fraudsters.
It takes highly effective and layered payment security to mitigate the complex threats of 2025. Behind layers of SaaS defenses, directed by expert human analysts, it’s a cyberwar ‘situation room.’
In its “2025 Fraud Trends” survey and white paper, financial crime risk management expert Themis sheds light on the state of banking fraud and the industry's preparedness to combat emerging threats. According to Themis, 50% of our responding banks “experienced an increase in fraud cases in 2024” with 17% reporting a decrease in fraud.
What has a lot of executives rattled are unfamiliar kinds of scams powered by generative artificial intelligence (GenAI), namely deepfakes and synthetic identities. If either of those attack types sound futuristic or unthreatening, you haven’t been following developments.
Digital Onboarding: A Double-Edged Sword
For starters, Themis said two-thirds of respondents (62%) identified online account opening and client onboarding as the top digital banking trends increasing fraud exposure.
That stat underscores the delicate balance between providing a seamless UX and maintaining robust security. For digital and mobile business payments, it’s safety first.
"New technologies like open banking and automation can initially seem safer, but they also create new opportunities for fraudsters,” said Bottomline risk expert Nick Griffin. “Banks need to be right every time to prevent fraud. Fraudsters only need to succeed once."
Odds do seem to favor fraudsters in the short term. That serves as a warning to banks, FIs, and corporates to rethink strategies now – and likely redouble digital investments – to shorten the cyberfraud dark ages and stop the losses.
Themis said 40% of survey respondents fear that synthetic identity fraud (SIF) “will pose one of the greatest threats to their bank in 2025…due to the complexities it presents for detection[,]” citing a Thomson Reuters analysis concluding that “95% of synthetic identities are not detected during the onboarding process at financial institutions.”
GenAI is Both Friend and Foe
Not surprisingly, Generative AI is now seen as a major disruptor, and the rules are being rewritten on the fly, even for enterprises. Themis said nearly half of respondents (48%) flag Gen AI as a significant force influencing fraud patterns, for good and ill.
"While AI can enhance our detection capabilities, it also equips fraudsters with sophisticated tools for creating deepfakes and synthetic identities,” Griffin said, echoing the study, He added that “The recent $25 million fraud in Hong Kong involving a deepfake illustrates the potential risks."
Griffin refers to the February 2024 incident in which a Hong Kong-based employee of a UK firm attended a video call with “company executives” during which that individual was instructed to transfer USD 25 million. Every executive on that call except the victim was a deepfake recreation, and convincing enough to be believed.
That was about a year ago. Deepfakes have now become even harder to detect.
File it under ‘putting out fire with gasoline’ but a more attuned GenAI is the tool best able to scan for anomalies, including deepfakes and synthetic identities, faster and more effectively than any technology now available. That makes GenAI both friend and foe to FIs and banks. The idea will take some getting used to as the technology becomes BAU.
Fraudsters in Our Midst
Insider fraud also gets a going over in the Themis report. This type of fraud, particularly challenging to detect, is on the rise as well. "Insider fraud has been underappreciated and underfunded in the past," Griffin said. "The blurred lines between insider and external fraud make it increasingly difficult to address. It's crucial to attack fraud from both internal and external perspectives."
Perhaps one of the most alarming findings is the number of banks feeling unprepared to address existing and emerging fraud risks. Griffin attributes this to the rapid pace of technological innovation: "The speed at which new technologies like generative AI and blockchain have become mainstream has left many institutions struggling to keep up,” he said. “This gap in preparedness is a significant concern that needs immediate attention."
There’s statistical support to back it up. Themis said 62% of respondents “find their current fraud prevention and detection technologies only ‘somewhat’ effective at addressing insider and external fraud risks simultaneously, which is a major gap in anti-fraud efforts.”
The Role of Closed-Loop Networks
Closed-loop B2B payment networks offer one solution to mitigate key fraud risks. Members are vetted and many find trading partners already present. "These networks create a trust ecosystem. That can significantly reduce the risk of fraud,” Griffin said, suggesting that the vetting process needs constant attention and innovation.
Combined with other significant trends, the “2025 Fraud Trends” report from Themis shows business banking at a crossroads.
Digital innovation brings many benefits, but it also introduces new vulnerabilities that organized and funded fraudsters can exploit. That’s their 9 to 5, and “they” are global. To stay ahead, banks must invest in not simply better fraud prevention tech, but also strategies and milestones that foster inter-organizational cooperation.
Adaptability and vigilance will be key to safeguarding financial institutions and their customers. By embracing these challenges and leveraging emerging technologies responsibly, the banking industry can build a more secure future for digital offerings.