Fraud Without Borders: Lessons from the Marketplace Risk Management Conference

By Adam Cail, Director of Enterprise Sales, Modulate

Ask anyone in risk management or fraud investigation and prevention and they’ll tell you: Fraud doesn’t stay in its lane. Fraud is a persistent issue in just about any industry, and fraudsters are often able to replicate their tactics across verticals with ease and efficiency. One tactic to claim charge-backs in gig delivery settings or online retail marketplaces can quickly migrate to banking, insurance, or even hiring platforms. Fraudsters are agile, collaborative, and increasingly armed with inexpensive AI tools that let them scale faster than traditional defenses can adapt. 

At this year’s annual Marketplace Risk Management Conference, leaders from across sectors emphasized one central truth: if you think fraud is “someone else’s problem,” it is only a matter of time before it reaches your business. Here are some of the key themes and learnings from this year’s conference.

Conference Insights

From panel discussions to case study analyses, this year’s MRM revealed just how industrialized fraud has become across different verticals: 

• Device farms and emulator markets are allowing automation at scale, with reports of up to 15,000 calls per day from a single device. Kristin Kupiec at DoorDash shared how her Critical Investigations team has to collaborate across device, account takeover, and location experts to track these evolving threats.

• PIN brute forcing and rapid resets make fraudulent behavior harder to detect and easier for bad actors to scale. As Danny Stalker from Incognia highlighted, fraudsters iterate very quickly on techniques like PIN testing, making containment extremely difficult.

• Fake applicant pools and synthetic identities are flooding hiring systems. During the session on candidate fraud, Karen from Persona and Josh from Yardstick explained that fake IDs and fraudulent applications are becoming easier to create as personal data becomes more accessible.

• Card scans, step-up challenges, and social engineering exploits continue to test delivery and payments systems. Travis Stine, also from DoorDash described their friction stack, which includes detection, enforcement, and resolution, and noted that even with safeguards in place, social engineering remains extremely powerful.

Different industries are seeing the same underlying problem: low barriers to fraud that have far reaching impact.

The Challenge

The biggest obstacle is certainly not the lack of effort. One key message throughout all the discussions this year was the persistence of silos. Fraud detection teams often split their focus by channel, device, or fraud method such as account takeovers. But these silos create a vulnerability. A fraud ring may begin with onboarding fraud in a hiring system, then pivot to account takeovers in delivery apps, and eventually apply the same playbook in financial services. As multiple speakers at the conference emphasized, no single person or team has all of the data needed to stop these attacks. When data, people, and systems remain siloed, fraudsters exploit the gaps.

At Modulate, we’re honing in on fraudsters that might spoof a device or fake an ID, and alter or fully synthesize their voice as they interact with call center agents and help desk reps, or even hiring teams. Whether it is a “customer” disputing a charge, an “applicant” onboarding with fake credentials, or a scammer testing stolen PINs, we find that real-time conversational analysis can reveal stress, evasiveness, or scripted behavior that metadata or deepfake detection alone cannot capture. Voice fraud detection is not just another point solution. It connects signals across industries, making it harder for fraudsters to migrate unchecked from one platform to another.

What’s Next in Fighting Fraud? 

Fraud prevention cannot remain siloed. The same fraud rings that are hitting delivery apps today will be targeting banking systems tomorrow. Fake hires in marketplaces will eventually test the defenses of support centers and retail CX. I see voice is a universal signal that surfaces intent in real time across all these use cases.

As Kristin at DoorDash said in her session, teams need to paint the entire picture and share it across functions to make progress. Modulate’s voice intelligence solutions give organizations a way to tell that story clearly and to act on it quickly. If you want to stay ahead of fraud that crosses industries, you need tools that are equally adaptable and borderless.