As AI creates opportunities to innovate banking services while enabling increasingly sophisticated cyberattacks and fraud, securing digital transactions requires an approach that goes beyond individual security systems. At Smart Banking 2026, Mr. Nguyen Xuan Viet, Chief Technology Officer and Director of the New Technology Services Division at FPT IS, FPT Corporation, proposed integrating cybersecurity monitoring, fraud detection, and anti-money laundering systems to strengthen real-time risk detection and response capabilities.
Panel discussion under Session 2, “Trusted Digital Banking: Driving Payment Innovation and Strengthening Fraud Prevention,” at Smart Banking 2026.The 7th Smart Banking Conference and Exhibition 2026 was jointly organized by the Vietnam Banks Association and IEC Group, under the auspices of the State Bank of Vietnam. The event focused on AI, customer trust, and digital resilience in the banking sector.
Speaking at the event, Mr. Pham Tien Dung, Deputy Governor of the State Bank of Vietnam, emphasized that AI adoption must be closely linked to practical effectiveness, risk governance, and security.
As part of the event, the panel discussion titled “Trusted Digital Banking: Driving Payment Innovation and Strengthening Fraud Prevention” brought together representatives from FPT, banks, and technology companies to address the challenge of balancing seamless payment experiences with the need to protect digital transactions.
Mr. Nguyen Xuan Viet of FPT Corporation highlighted two major categories of risks that banks need to prioritize: direct cyberattacks targeting technology systems and fraud targeting customers through impersonation and psychological manipulation.
According to Mr. Viet, these two categories are becoming increasingly difficult to separate, as many cyber intrusions now begin by deceiving individuals before attackers gain access to systems using legitimate login credentials.
Mr. Nguyen Xuan Viet of FPT shares insights during the panel discussion on driving payment innovation and strengthening fraud prevention in digital banking.According to Mr. Nguyen Xuan Viet, the first challenge is that AI is making cyberattacks faster and more sophisticated.
While banks are adopting AI to enhance customer experiences, optimize processes, and develop digital services, attackers can also exploit the technology to automate vulnerability discovery, accelerate intrusions, and scale up attacks.
This requires banks to move from a reactive defense approach toward continuous monitoring, early threat detection, and proactive incident response.
Security management must also extend beyond human users to include machines, applications, and AI agents capable of autonomously performing assigned tasks.
The second challenge involves AI-generated deepfakes and psychological manipulation, which create risks that conventional transaction controls may struggle to detect.
Fraudsters can use AI-generated or manipulated images, videos, and voices to impersonate real individuals and deceive customers or bank employees, potentially persuading victims to initiate money transfers themselves.
In such cases, transactions may still pass conventional authentication checks if banks verify only login credentials or identity without examining unusual behavioral patterns or the circumstances surrounding a transaction.
Banks therefore need to combine multiple authentication layers, transaction behavior analysis, warning intelligence, and coordination with relevant organizations to identify potential fraud at an early stage.
In response to these challenges, the FPT representative proposed three approaches to strengthening digital banking security.
1. Strengthening Security Foundations and Identity Management
Banks need to upgrade their security capabilities based on the Zero Trust principle, which assumes that no user or device should be trusted by default. Every access request must be verified based on identity, authorization, and the specific context.
Zero Trust should apply not only to banking infrastructure and employees but also to customers and individual transactions.
This includes linking accounts to registered devices and adopting passwordless authentication based on the FIDO2 standard, which uses cryptographic technology to verify user identity and enables authentication through fingerprints, facial recognition, or device PINs instead of passwords.
Banks must also ensure that the transaction details confirmed by customers match the actual transaction being executed.
This approach helps reduce the risk of unauthorized changes to payment instructions during the confirmation process.
To address AI-powered impersonation, the FPT representative also recommended exploring additional capabilities for Injection Attack Detection (IAD) alongside the already mandated Presentation Attack Detection (PAD) measures.
These capabilities would strengthen biometric authentication systems against emerging attack techniques.
2. Integrating Cybersecurity and Fraud Prevention Systems for Real-Time Risk Detection
According to Mr. Viet, banks currently operate multiple specialized systems, including:
- Security Operations Centers (SOC): Monitor technology infrastructure and detect signs of cyber intrusion.
- Fraud Detection systems: Identify suspicious transactions and unusual transaction behavior.
- Anti-Money Laundering (AML) systems: Help detect transactions potentially associated with money laundering activities.
However, when these systems operate independently, each may capture only part of an incident. As a result, banks risk overlooking connections between cyberattack indicators, unusual login behavior, and suspicious transactions.
Integrating signals and data from these systems would enable banks to develop a more comprehensive view of risk, correlate related indicators, and prioritize high-risk situations.
Rather than evaluating alerts individually, banks could identify sequences of suspicious activities within the same context, from intrusion attempts and abnormal logins to questionable financial transactions.
This would shorten the time required to detect and respond to cyberattacks, scams, and fraudulent activities.
3. Applying AI to Cybersecurity Monitoring and Incident Response with Appropriate Controls
FPT is developing an AI-powered SOC approach to support security alert analysis, investigation, and incident response.
AI can process large volumes of data, identify significant indicators, and recommend response actions, enabling cybersecurity specialists to make faster decisions.
Under this approach, AI should not only accelerate operational workflows but also ensure reliability in analysis and recommendations.
Actions with potentially significant consequences must remain subject to appropriate control mechanisms, including expert supervision and approval when necessary.
With 30 years of experience supporting Vietnam's banking and financial sector through digital transformation, FPT currently serves more than 100 domestic and international banks and financial institutions through a comprehensive technology ecosystem encompassing AI, cloud computing, AI infrastructure, cybersecurity, intelligent automation, and data governance.
A delegation led by Mr. Pham Tien Dung, Deputy Governor of the State Bank of Vietnam, visits the Made by FPT technology demonstration area.FPT's Made by FPT solutions address a wide range of banking operational needs, reflecting FPT's strategy of developing and mastering core technologies.
AI-Powered SOC is an AI agent-powered security operations center platform that supports security monitoring and threat detection across information systems.
Akabot provides an intelligent automation ecosystem designed to optimize business processes.
TradeFlat supports the digitalization of trade finance and supply chain finance operations.
Meanwhile, Data Governance solutions help banks manage, protect, and effectively unlock the value of their data.