Battle of the Bots: 3 Key AI Insights From Money20/20 Asia

The intensifying battle between AI-powered fraud and AI-driven defenses took center stage during a panel discussion at Money20/20 Asia.

The “Battle of the Bots” panel included leaders from Grab, AXS, DiDi and Trulioo. They discussed how financial services and marketplace enterprises can sharpen their defense strategies as fraud becomes more advanced. 

Here are three key insights from the session that can help global enterprises outmaneuver even the most sophisticated bad actors.

1. Local Precision Meets Global Scale

AI has lowered the technical barriers for bad actors, making sophisticated technology accessible to anyone. Attack strategies that once required deep expertise are now widely available through tools sold on the dark web, complete with how-to guides. 

Reactive solutions are no longer enough, said Puneet Gambhir, Grab’s head of risk and GrabDefence Business. Enterprises now can rely on dynamic, scalable fraud detection programs that combine data standardization with local expertise. 

“The infrastructure and the code solution need to be very standardized,” he said, “but the intelligence that you build on top of it needs to be very local.” 

Enterprises that can combat fraud at scale while adapting to regional nuances can build a strong, lasting defense.

2. Human Oversight Remains Irreplaceable

AI can process vast amounts of data and flag suspicious behavior, but only humans can provide context and assess intent.

“AI helps identify the anomalies, but humans connect the dots,” said Caitlin Wanlin Li, DiDi’s senior AML strategy analyst. 

False positives are common in areas such as anti-money laundering. Human fraud experts can play a crucial role in piecing together user behavior, identity data and transaction history to determine if an alert warrants further action. 

“AI is only effective if the data is clean, and if humans are there to validate and act on it,” said Sean Tan, AXS head of risk and compliance. 

Rather than replace human oversight, AI shifts the role of risk professionals. Those experts now supervise automated systems, acting as gatekeepers who can validate and refine AI-driven decisions.

3. Collaboration Is Key

Fraudsters don’t respect borders or business lines, and neither should those who defend against it. 

The panelists agreed that data sharing and industrywide collaboration are essential in outpacing fraud’s rapid evolution. 

“In the future, we’re not only fighting AI with AI,” Wanlin Li said. “It’s fighting fragmentation with collaboration.”

From joint AI sandboxes to government data-sharing agreements, regional and cross-sector partnerships can help seal intelligence gaps and align with regulatory standards. 

Still, challenges remain. Gambhir emphasized that data must be shared in actionable formats that reflect how modern fraud correlates to user behavior patterns. 

“Yes, let’s collaborate,” he said, “but collaborate to do what?”

Powerful Fraud Defenses Require the Complete Toolbox

Beating next-generation fraud takes more than advanced technology. It requires adaptive processes, empowered human oversight and strong collaboration across stakeholders. 

AI may be the weapon, but strategy, coordination and human judgment will win the war.

“It really takes a whole village to come together,” Tan said, “and get this going from a collaborative angle.”