Thailand’s online scam epidemic has become far too large and sophisticated to be simply treated as a law-enforcement issue. As more consumers conduct transactions using digital channels, fraudsters are exploiting these exact systems that have made banking services convenient and quick. Social engineering tactics, online payment systems, false identities and mule accounts make it possible for criminals to move stolen funds quickly through the financial system before victims or institutions have time to react.
The scale of the problem is significant. Thailand recorded more than 363,000 online fraud cases and around THB24.57 billion in losses in 2025, underlining how deeply digital fraud has become embedded in the country’s online economy. The number of complaints handled by Thailand’s Electronic Transactions Development Agency also reached 39,112 cases in 2025, showing that online problems extend beyond conventional financial fraud to a broader range of digital harms.
We take a look at why Southeast Asia faces a wave of forced tech exits over the next year
The speed of fraud is changing the challenge
Speed has become one of the defining features of online fraud. Fraudsters no longer necessarily need to steal money through a sophisticated technical breach. Instead, they manipulate people into authorising transactions themselves and then transfer the money rapidly from one account to another. The central bank of Thailand has highlighted the increasing sophistication of financial fraud. This includes social engineering and the use of mule accounts to move funds. As soon as the money enters the system, criminals can move it through multiple accounts, making recovery considerably more difficult.
The timing gap between fraud and reporting makes this particularly difficult. Research highlighted in Thailand’s scam response efforts suggests fraudsters can move 50 per cent of stolen funds in an average of three minutes, while victims may take hours to report what has happened. By the time a bank or law-enforcement agency receives a report, the money may already have passed through several accounts or been converted into other assets. Fraud happens in minutes, while institutional responses traditionally happen much more slowly. Thailand is therefore building an increasingly real-time, data-sharing fraud infrastructure across banks, regulators, telecoms and digital platforms. AI can become one component of that infrastructure.
AI could move fraud detection upstream
Traditional anti-fraud systems have often relied heavily on fixed rules, transaction thresholds and known indicators of suspicious activity. But these approaches can struggle with scam transactions that appear legitimate because they are authorised by genuine customers. AI-assisted and real-time monitoring can potentially assess a broader range of behavioural signals before funds move further through the system.
This is particularly important in cases involving mule accounts. An account that suddenly starts receiving multiple transfers, rapidly moves money elsewhere and operates very differently from its previous activity may present a risk that is difficult to identify through static rules alone. Thailand is already moving towards more data-driven fraud prevention, including cross-bank information sharing through the Central Fraud Registry, stronger monitoring of abnormal account behaviour and tighter controls on suspected mule accounts. The Bank of Thailand has adopted certain measures aimed at making financial institutions accountable for fraud risks and strengthening monitoring of suspicious activity.
Identity verification is becoming a frontline defence
The other significant dimension is identity. With the evolution of scammers’ techniques to pose as genuine institutions and people, understanding the actual identity of those running the accounts has become more relevant than ever. More rigorous KYC policies, biometric identification, behavioural analysis and device intelligence may complicate the process of generating or using scam accounts. Nonetheless, the identity verification process will not be effective if it is carried out exclusively at the stage of opening an account. Businesses increasingly need continuous monitoring that can identify when an account’s behaviour changes significantly after being onboarded.
This is particularly relevant to mule accounts. Criminal networks can recruit individuals to open accounts legitimately before using those accounts to receive and transfer illicit funds. Detecting this behaviour requires financial institutions to look beyond the identity document submitted at registration and understand how an account is actually being used. The challenge is therefore moving from static identity checks towards dynamic identity and risk assessment.
Thailand is tightening controls across the ecosystem
Thailand has also strengthened its regulatory and operational response across the financial system, with measures increasingly focused on disrupting scams before money moves beyond recovery. This stems from the recognition that online scams are spread out across multiple levels in the digital economy. For example, a scam could start off with an advertisement on a social media platform, proceed to a messaging app, include a fraudulent website and finally conclude with a transfer through a bank or digital payment service provider.
This means there is no single organisation overseeing the whole process. Hence, the need for cooperation increases. Cooperation among financial institutions, telecommunication companies, online platforms and law enforcement is one of the measures the Thai government has been implementing as part of the broader effort to address online fraud.
Fraud prevention could create opportunities for technology providers
Fraud intelligence, transaction monitoring, identity verification, behaviour analytics and digital investigation technologies have increasingly become relevant as financial transactions have transitioned online. This also creates potential opportunities for startups and technology providers because fraud prevention is relevant well beyond banking. Banks and payments companies will require fraud detection services, while e-commerce websites, telecom companies, insurance firms and online marketplaces can also face risks linked to fraudulent accounts and impersonation.
Nevertheless, technology alone is unlikely to be able to resolve the matter at hand. An advanced fraud detection system will not work well if an organisation does not have the necessary processes, manpower and regulatory framework required to act on its warnings. The strongest response is therefore likely to combine artificial intelligence with human investigation, automated risk scoring with clear escalation procedures and identity technology with stronger institutional cooperation.
Thailand needs to make scams harder
Thailand’s online scam epidemic reflects a broader challenge facing digital economies across Southeast Asia. As digital payments and online services become more deeply embedded in everyday life, criminals will continue looking for weaknesses between financial institutions, platforms, telecommunications networks and consumers. The response therefore needs to evolve at the same speed.
Thailand has already begun moving fraud prevention towards a more proactive model. The next challenge is to make those systems faster, more predictive and better connected across institutions. AI-assisted detection, behavioural analytics, stronger identity verification and real-time transaction monitoring could push intervention even closer to the moment suspicious activity begins.
Thailand’s experience highlights a broader opportunity for Southeast Asia in cybersecurity and fraud prevention as these become a vital part of the infrastructure of the digital economy. The ability of banks, platforms, telecommunications companies and governments to collaborate quickly enough to identify and disrupt suspicious activity before funds move beyond recovery may not only reduce losses, but also build greater trust in the next generation of digital services.