AI-era fraud pushes digital finance from transaction blocking to user protection

AI-era fraud pushes digital finance from transaction blocking to user protection

As AI-powered scams grow more sophisticated, banks and digital platforms are shifting from simply blocking fraudulent transactions to protecting user behaviour and trust in the digital financial system.

Customer experiences MoMo’s digital financial services and anti-fraud features at the Digital Trust in Finance 2026 forum in Hà Nội. — Photo courtesy of MoMo

Data from the Ministry of Public Security shows Vietnamese users lost an estimated VNĐ8 trillion (US$304 million) to online scams in 2025 alone. However, experts warn that the damage goes far beyond financial losses and threatens public trust in the country’s rapidly expanding digital economy.

The rise of generative AI is making fraud schemes cheaper, faster and more convincing. Tasks that once required advanced technical skills can now be carried out using widely available AI tools capable of generating fake images, videos, voices and financial data.

Instead of targeting banking infrastructure directly, scammers are increasingly exploiting fear, urgency and the lure of quick profits to persuade victims to transfer money voluntarily.

One common scam model in Việt Nam involves fake AI-powered investment platforms promoted through social media and online chat groups. Scammers use professional-looking apps, fabricated profit data and self-styled investment mentors to lure victims into depositing money. They often allow small withdrawals at first to build trust before pushing larger investments.

“The problem today is not simply fraudulent transactions. In many cases, users themselves authorise the payments while being manipulated psychologically,” Nguyễn Mạnh Tường, co-founder and chief executive of fintech platform MoMo, said at a recent forum on digital trust in finance. 

Customer experiences MoMo’s digital financial services and anti-fraud features at the Digital Trust in Finance 2026 forum in Hà Nội. — Photo courtesy of MoMo

Experts say the evolving nature of online fraud is forcing financial institutions to rethink traditional cybersecurity strategies.

For years, most fraud prevention systems focused mainly on detecting suspicious transactions after they occurred. But many scams now succeed not because banking systems are breached but because victims are manipulated into making transfers themselves.

As a result, banks and digital finance platforms are shifting from blocking suspicious transactions towards using systems to detect behavioural risks and intervene before users make irreversible decisions.

According to MoMo, timely risk warnings can significantly reduce fraud losses, with many users pausing transactions to verify information after receiving alerts.

Tường said this reflects a broader shift in cybersecurity thinking from protecting systems alone to protecting user behaviour and decision-making.

To strengthen fraud prevention, MoMo has developed a multi-layered AI-based defence system that analyses transaction risks, detects biometric fraud such as deepfakes and identifies suspicious accounts and scam networks.

The company says the system detects nearly 29,000 abnormal transactions and warns more than 10,000 users about potential risks each day, helping protect more than VNĐ15 trillion annually from fraud each year.

However, Tường warned that individual financial institutions cannot solve the problem alone, as scams often unfold across multiple platforms before funds are transferred through banks or e-wallets.

Many schemes now begin on social media or messaging apps, where scammers build trust with victims over days or even weeks before directing them to make financial transfers.

Meta recently said it removed more than 159 million scam advertisements worldwide in 2025 for policy violations, while disabling 10.9 million Facebook and Instagram accounts linked to organised scam centres.

The company also said it participated in joint operations with international law enforcement agencies that led to the removal of more than 150,000 scam-linked accounts and assisted Thailand's authorities in making arrests.

Southeast Asia has become a major hub for organised cyber-enabled fraud networks in recent years, according to multiple international investigations, with criminal groups increasingly operating across borders and targeting victims through digital platforms.

Vietnamese authorities are now calling for stronger coordination between banks, telecom providers, social media platforms and regulators to improve early fraud detection and speed up responses before stolen funds disappear.

Nguyễn Hồng Quân, deputy director of the Ministry of Public Security’s Department of Cybersecurity and High-Tech Crime Prevention, said faster information-sharing mechanisms are needed so suspicious accounts and fraudulent transactions can be identified and blocked more quickly.

Quân also stressed the importance of stronger identity verification systems and better protection of official digital communication channels to help users distinguish legitimate platforms from fake ones.

He added that personal data protection should be treated as a core pillar of digital financial trust, with banks and financial service providers expected to take greater responsibility for safeguarding user information.

As AI continues to lower the barriers to cybercrime, experts say the challenge for governments, financial institutions and technology platforms is no longer simply preventing fraud, but maintaining public trust in the digital financial system and the wider digital economy. 

Bizhub