Online scam syndicates that rely on trafficked workers and guarded compounds could soon run fraud operations autonomously using agentic AI, according to a cybercrime expert. The shift threatens to make scams faster, cheaper, more personalised and harder for authorities to trace.
Losses from transnational online scams in Asia-Pacific tripled to US$114 billion last year from 2023, according to a United Nations report released last month. The surge reflects an "increasingly integrated, transnational, and technologically sophisticated criminal economy" that shares expertise in money laundering, human trafficking and data harvesting.
Jonno Newman, a cybercrime expert with on-chain intelligence platform TRM Labs, said the complexity of scam operations in the region could grow exponentially once criminals master agentic AI. Unlike earlier AI uses for writing scripts, translating messages, or producing deepfakes, agentic systems can execute multi-step tasks with minimal human prompting.
How agentic AI changes the scam economics
Newman said syndicates were already investing heavily in training agentic AI agents to work with years of data compiled by their large language model predecessors, allowing them to target multiple victims across different languages simultaneously.
"So instead of one human, they can use a bot that can do the work of 10 people. You would no longer need a big fortress - a scam operation could be set up from anywhere remotely," Newman told This Week in Asia.
This collapse of scale matters for detection. Traditional scam infrastructure - guarded compounds, call centres, large workforces - gives law enforcement physical targets. Autonomous operations could be run remotely with far fewer people involved, shrinking the footprint that investigators rely on.
Faster and harder to trace
The practicality of agentic AI lies in its ability to combine decades of conversation data with real-time personalisation across languages. A single operator could run thousands of interactions at once, adapting each conversation based on victim responses without human review.
Sensitive: scams relying on trafficked workers will not disappear. The UN reported that this criminal economy has been integrating shared expertise in money laundering and human trafficking. But if agentic AI reduces the need for massive call centres, the human-trafficking element and the physical scale of current fraud compounds may shrink - while the volume of scam attempts could continue to rise.
Why this matters for operations professionals
For people working in operations across Asia-Pacific, the threat is concrete. Fraud has already tripled in the region, and your supply chain or vendor network could easily touch the tech that drives it - from payment rails that inadvertently move scam money to customer support systems that mimic AI-generated dialogue.
If agentic scams become the default, operations teams should expect strange changes: more personalised phishing attempts, higher-quality fraud at lower, and contacts asking for payments in smaller, staggered increments to dodge suspicious-activity flags. That means updating vendor review, strengthening account verification, and building in redundancy checks that catch anomalies before transactions clear.
Your membership also unlocks:
