AI scams escalate as criminal networks use cheap tools to evade crackdown and dupe more victims

Criminal Scam Centers Embrace AI to Scale Operations

Criminals in Southeast Asia are harnessing inexpensive AI tools to transform the global scam economy, targeting bigger pools of potential victims at high speed while staying one step ahead of law enforcement. According to senior officials at Interpol, scam centers continue operating at full capacity despite government crackdowns, thanks to artificial intelligence that makes their operations cheaper, faster, and far more convincing 1

1

.

"You can see the efficiency with AI being utilized in scam centers," said Neal Jetton, who leads the Cybercrime Directorate at Interpol in Singapore. "It's a pretty easy business model, and I think it's going to get even easier for criminals with AI" 1

1

. The shift represents a fundamental change in how transnational criminal networks operate, with AI scams becoming increasingly difficult to identify.

How Large Language Models Transform Scam Operations

The biggest transformation involves speed and sophistication. Previously, scams were relatively easy to spot through poor quality online ads or grammatically flawed messages. Now, scammers deploy large language models like ChatGPT to craft authentic-sounding messages in multiple languages, even those they don't speak fluently 2

2

. Stephanie Baroud, a criminal intelligence analyst with the human trafficking and migrant smuggling unit at Interpol in Lyon, France, explained that realistic-looking ads can now be generated within seconds using the right prompts 1

1

.

Voice cloning and deepfake technology add another layer of deception. Scammers use voice impersonation calls to pretend to be someone's child or grandchild, creating emotional pressure that makes victims more likely to send money. Video impersonation through deepfakes allows men to wear digital masks and appear as attractive women, helping them dupe more victims across the world 1

1

. AI also enables quick image generation for fake profiles, making romance and investment scams far more believable.

Financial Losses Reach Staggering Levels

The financial impact of AI-powered cybercrime is enormous and growing. A US Institute of Peace report estimates that by end-2023, transnational criminal networks operating illegal online gambling and scam operations stole approximately $64 billion annually worldwide 1

1

. This figure is likely conservative, as many victims never report being scammed. The true cost to society remains difficult to quantify, but experts agree it continues rising as AI makes operations more efficient and scalable.

The low cost of AI tools means criminal operations can afford to take more risks. Simply put, AI allows operators to scale up at minimal expense, making them more willing to run the risk of being caught and quickly relocate when law enforcement closes in 1

1

.

Scam Landscape Expands Beyond Southeast Asia

While Southeast Asia remains the epicenter, the scam landscape is globalizing rapidly. Interpol's Baroud noted that more operations are appearing in the Americas, Africa, and the Middle East. Some maintain Asian elements suggesting links to gangs in Cambodia and Myanmar, while others are run by local organized groups, indicating the globalization of the modus operandi itself 1

1

.

Recent crackdowns in Cambodia, following pressure from the US, China, and Thailand, have led to high-profile arrests. Chen Zhi, an alleged kingpin of an international syndicate facing charges in the US, was arrested in Cambodia and deported to China. Beijing also announced it had executed 11 people linked to scam centers in Myanmar 1

1

. Yet these actions haven't significantly reduced operations.

Law Enforcement Struggles to Keep Pace

Julia Dickson, an associate fellow in the intelligence, national security and technology program at the Center for Strategic and International Studies, warns that "the industry's just become increasingly professionalized, which is concerning obviously with the sort of explosion of AI and the availability of it" 1

1

. While AI has potential to help law enforcement track fraudsters, authorities are struggling to match the pace at which criminals adopt new technologies 2

2

.

Baroud emphasized that AI is reshaping the phenomenon across all actors involved: "Whether by the organized groups themselves, by the potential trafficked victims, or by the people being scammed, it is likely reshaping the phenomenon. AI is going to lead us into uncharted territory" 1

1

. Human trafficking into scam compounds hasn't decreased either, as operations still require workers despite technological advances 3

3

. The ability to rapidly rewrite scripts, switch languages, target new regions, and evade crackdown when authorities intervene makes these operations particularly resilient 2

2

.