How Jian Ming is tackling scams at their roots - Public Sector Transformation Awards 2026

It’s a simple analogy, but one that captures how Jian Ming approaches the fight against scams. For many people, anti-scam efforts feel like an endless game of whack-a-mole: take down one scam website and another quickly appears under a different domain; block one number and scammers simply switch to another SIM card. 

Jian Ming sees the problem differently. Rather than chasing one scam after another, he believes the real challenge is understanding how scam operations work and disrupting them at their source. Whether discussing machine learning or international collaboration, he consistently returns to the same principles: stop scams earlier, automate where it makes sense, and tackle the infrastructure that enables scams rather than simply reacting to them. 

Meet Jian Ming, Technical Lead at GovTech's Anti-Scam Products (GASP) team, driving three interconnected AI/ML innovations at the heart of Singapore's fight against online scams. This year, he received the Exemplary Innovation Award at the Public Sector Transformation Awards 2026. 

In this interview, Jian Ming shares how he and his team moved beyond reactive scam takedowns to build a closed-loop system that identifies, understands, and disrupts threats. He also highlights how the team is taking Singapore's anti-scam capabilities global through the Global Signal Exchange, and what's next as the GASP team works to push its defences closer to citizens.

Staying a step ahead of a well-resourced opponent 

 There is no routine day for Jian Ming. Some days are spent troubleshooting a pipeline issue or responding to a newly surfaced scam variant; others are spent working closely with partners to coordinate the next response. 

One thing, however, remains constant.

This mindset matters because Jian Ming is quick to correct a common misconception that scam operations are small-time, opportunistic, or amateurish. 

"These aren't lone actors working out of a bedroom. They're organised, well-resourced, and constantly iterating," he explains. "We aren't just building something that works today. We're building something that keeps learning alongside a very motivated adversary." 

From detection to disruption, closing the loop 

Before Jian Ming's innovations, stopping a scam relied heavily on human judgement, with analysts reviewing reports, cross-checking indicators and making decisions on a case-by-case basis. 

That worked at low volumes, but scam operations have since scaled enormously. 

The team's three interconnected innovations didn't start out as a single master plan. As they mapped the scam lifecycle from identification to disruption, it became clear that addressing any one stage in isolation wouldn't be enough. 

"The idea was to bridge the gaps between identification, analysis, and disruption," Jian Ming explained. “The three innovations naturally came together out of necessity to create a closed-loop system that identifies threats, understands them, and takes them down almost simultaneously." 

 
Together, the three innovations work as one integrated system: 

• Fan Out: For every malicious site the system finds, it traces that site's "friends", shared assets, page titles, domains, and registrants to uncover clusters of related scam sites from a single lead. 

• Feeds Unlimited: The automated factory that handles the volume “Fan Out” creates. Every incoming feed runs through rigorous classification before triggering takedown at the browser level. 

• AI/ML Blocking: For the highest-confidence cases, the system can trigger takedown and issue legal orders automatically, work that used to require a police officer's manual review. For the lower-confidence cases, a built-in safety net re-routes cases back to human analysts, balancing speed with accuracy. 

Going closer to the source 

For Jian Ming, tackling scams at the root means working further upstream, and often beyond Singapore's borders. 

"If not, we'll just be playing whack-a-mole. Every time a scam website comes out, we just take it out. But the people behind it are still out there,” he shared. 

That thinking has shaped the GASP team's approach to disrupting scam infrastructure instead of simply taking down individual scam websites. 

One example was when the GASP team's data showed that a disproportionate share of scam domains traced back to a single Singapore-based domain registrar. Rather than continuing to take down fraudulent websites one by one, the team worked directly with the registrar to tighten Know-Your-Customer (KYC) checks at the point of registration. 

As a result, the proportion of scam domains traced back to that registrar fell from around 25% to just 5%, demonstrating how solving the problem closer to its source can deliver measurable results. 

That same philosophy also extends beyond Singapore. The GASP team is among the first government teams globally to join the Global Signal Exchange (GSE), sharing signals in real time with international law enforcement and major technology companies. The team is also working with partners across ASEAN to strengthen regional collaboration, recognising that scam infrastructure often operates across borders. 

Every scam prevented is a win 

The three innovations have since earned Jian Ming and his team the Exemplary Innovation Award, recognition he describes in team terms rather than individual ones. 

For Jian Ming, there is more meaning behind the work than the award itself. 

"It feels great! When you see the numbers, from scam sites taken down to potential victims protected, it reconnects you to why you're doing this. Our metrics represent real people who didn't lose their life savings," he said. 

The scale of success for anti-scam efforts is hard to see, simply because the system is working, and there is no easy way to show a citizen what didn't happen to them. 

Finding a perfect balance between caution and aggression 

When asked about the hardest part of the job, Jian Ming doesn't point to the scammers first. He points to getting the balance right.  

Being too aggressive generates false positives that erode trust in the system and its partners. Being too cautious allows real scams to slip through. 

As scam tactics continue to evolve, that balance must be constantly recalibrated to ensure the system remains both accurate and effective. 

What’s next for the anti-scam scene 

"A major focus right now is tackling the platforms where scams originate from, specifically malicious social media ads and encrypted messaging apps. To effectively protect users on these apps, we are exploring device-level protection against scams. The threat landscape is evolving quickly, and our goal is to push our defenses as close to the user as possible."  Jian Ming shared. 

It is a quiet extension of what GovTech has long worked towards: infrastructure that keeps Singapore's government, and by extension its people, secure.  

Such work is largely invisible by design, where systems like Jian Ming's have often already flagged the scam site, traced its network, and shut it down, actively defending people's digital lives, before they even notice a threat. 

For all the AI and infrastructure behind the work, Jian Ming's advice to citizens is simple: pause before you click. 

"Most scam sites are designed to create urgency: a limited-time offer, or a warning that your account is compromised. That pressure is intentional. It's meant to short-circuit your critical thinking. If you feel rushed, that is exactly the moment to slow down," he warned. 

It is a small habit against a large, shape-shifting problem. Yet it's the same instinct behind everything Jian Ming builds: don't just swing at the mole, go after what keeps sending them up. Whack-a-mole only ends when someone stops playing by its rules and for Jian Ming, his goal is simple: Making sure Singapore never has to play it in the first place. 

Connect with us 

Subscribe to the TechNews email newsletter 

Join the Tech Kaki user-testing community