RMIT expert calls for tougher tech response to scam threats

Crackdowns rise on scam compounds, more work still needed

In a major escalation against Southeast Asia’s scam networks, Meta has disabled more than 150,000 scam‑linked accounts following a coordinated enforcement sweep led by the Royal Thai Police, the FBI, and the US Department of Justice, with support from agencies in other countries. The operation, which follows a December pilot that removed 59,000 accounts, reflects growing urgency as scam compounds have been expanding their global reach and technological sophistication.

“Even though shutting down over 209,000 accounts may seem like a lot, we need to remember that Facebook has an estimated three billion users and malicious criminals can easily create more accounts,” Dr Joshua Dwight, a cyber security expert and associate program manager at RMIT University Vietnam, notes.

Meanwhile, Cambodian authorities announced they have shut down 80% of known scam‑centre locations and aim to eliminate the rest by April 2026. The country has been a major hotspot for industrial‑scale scam compounds, with operations expanding rapidly in recent years.

Against this shifting backdrop, Dr Dwight warns the digital ecosystem that enables scam compounds remains highly adaptive and far harder to eradicate than the physical compounds themselves. Therefore, continued awareness and action are needed to tackle this transnational crime.

How technology drives both recruitment and exploitation

Dr Dwight explains that scam compounds use technology in two different ways: first to recruit and traffic victims, and then to train and force them into cyber‑enabled scams once inside the compounds.

“Scam compounds use a wide variety of social engineering techniques to recruit, control, and traffic victims to areas such as Cambodia and Myanmar, where they are exploited,” Dr Dwight notes.

He highlights malvertising – fake digital ads on Facebook, Zalo, Telegram, Messenger, WeChat and other platforms – promoting high‑income or “easy work” opportunities as a primary lure.

“These fake ads or communications will offer high paying jobs, get‑rich‑quick gambling, crypto, and romance types of schemes,” he says, adding that even students and young people with strong digital skills have been targeted and fallen victim to these tactics.

Once trafficked, victims face a second wave of technological manipulation. “They forcibly train victims to use social engineering technologies such as deep fakes, VoIP impersonation, registration of malicious websites and domains that mimic real business sites to steal money from people globally,” Dr Dwight says.

Recent Cambodian raids also confirm this industrial model, with police seizing fake police IDs and uniforms used for impersonation scams, alongside evidence of cryptocurrency fraud operations.

Strengthening platform defences and user awareness

Dr Dwight says that social media platforms need to do better at verification of accounts and businesses that advertise on their platforms.

“They have not been incentivised to do much since they appear to gain quite a big revenue from malicious activities,” he says, citing reports that Meta showed users 15 billion scam ads per day, generating an estimated 10% of its revenue in 2024.

He emphasises that while Meta’s new tools – such as alerts for suspicious friend requests and WhatsApp warnings about fraudulent device‑linking – are a positive step, platforms remain overwhelmed by scale.

Therefore, continued investments in awareness campaigns are essential. “The more people are aware and trained to recognise scams, the less they will fall into these scam compound traps,” he says.

He suggests that social media platforms can also use anomaly detection and information sharing with organisations, law enforcement, and other non-governmental agencies to help identify and block emerging social engineering trends.

Anomaly detection could risk-rate online advertisements targeting Vietnamese citizens that use keywords from local law enforcement such as “lương cao” (high salary), “việc nhẹ” (easy work), and “tuyển gấp” (urgent recruitment).

AI agents within social media platforms can show “nudges” to warn people of potential scams or unknown numbers. “AI agents can be used to identify suspicious messages and links in advertisements, and warn the users to be careful about unsolicited job offers requiring relocation,” he says.

In addition, Dr Dwight says social media platforms and users could collaborate to share information with public data repositories and organisations such as VirusTotal, Viettel, the Vietnam Ministry of Public Security’s Department of Cybersecurity and High-Tech Crime Prevention, and ASEAN Regional Computer Emergency Response Team (CERT).

Ultimately, Dr Dwight believes the public must understand that many “scammers” are themselves victims. Cambodia has repatriated nearly 10,000 scam‑centre workers from 23 countries in the latest crackdown – many trafficked through the same digital deception they were later forced to inflict on others.

“I think that people often overlook the different creative ways that criminals implement technologies to conduct their malicious activities. Criminals are very good at finding vulnerabilities in people and processes and use digital tools very fast and very effectively. Vigilance is thus essential,” he says.

Story: Ngoc Hoang

Masthead image: biswas – stock.adobe.com | Thumbnail image: Kevin – stock.adobe.com