In early 2025, Block—the parent company of Cash App—agreed to pay up to $255m to settle claims that criminals had used its platform for scams and money laundering due to insufficient anti-fraud controls.
According to Resistant AI, unfortunately, this scenario is far from unique. Payment fraud has become a systemic threat, often beginning during onboarding when fraudsters use fake or stolen documents to gain access. Once inside, they exploit platform features to move funds rapidly, often routing them through money mule networks before detection.
The fraud doesn’t stop there—it spreads. Transaction systems become infected, requiring advanced detection tools like behaviour analysis and anomaly detection. To stop fraud at the source, platforms must adopt a layered defence model that integrates secure onboarding, document verification, and real-time transaction monitoring.
Payment fraud involves unauthorised or deceptive transactions carried out with manipulated or stolen credentials. The most effective way to counter this is through tools such as document fraud detection software from Resistant AI, which helps block bad actors before they can act.
It’s crucial to understand that payment fraud strikes at two key stages: onboarding and transaction. With many schemes overlapping with financial crime (fincrime), such as money laundering or terrorist financing, the lines between fraud and fincrime are increasingly blurred.
Fraud typically begins with access. Weak onboarding allows scammers to create synthetic identities or hijack real ones. If KYC and KYB controls fail, fraudulent accounts can easily exploit platforms. Platforms need more than surface-level checks—authentication must go deeper to assess document integrity and behavioural clues.
If these criminals slip past onboarding, transactional fraud comes next. Using stolen or synthetic credentials, fraudsters move funds quickly, often laundering money or evading sanctions. Real-time monitoring powered by AI can detect subtle anomalies that traditional rules-based systems miss. AI’s ability to connect disparate signals is key to spotting fraud as it happens.
In 2025, the fraud landscape is being shaped by new technologies and more aggressive tactics. Fraudsters mimic legitimate users, using AI tools and automation to bypass basic defences. Neobanks and newer platforms, due to rapid user growth, are particularly vulnerable. But even legacy institutions aren’t immune—fraud is simply more concentrated where onboarding is highest.
Most payment processors still rely on outdated rules or manual review, often identifying fraud too late. Meanwhile, P2P payment apps like Cash App and Zelle, designed for instant transfers, provide criminals with ideal channels for laundering funds. Their ease of use becomes a vulnerability.
APP fraud, synthetic identity abuse, and money mule scams are now driving staggering losses. APP fraud alone makes up 75% of all digital banking fraud, and with underreporting widespread, the actual figures are likely far higher. Mandatory reimbursement regulations like the UK’s new law are now holding platforms financially accountable.
This liability forces platforms to overhaul their defences. They are spinning up fraud teams, revisiting onboarding pipelines, and investing in compliance to avoid fines and reputational damage. Delays, false positives, and a heavier burden on legitimate users are byproducts of poorly implemented prevention strategies.
Common payment fraud tactics vary—from APP fraud and identity theft to chargeback abuse and account takeovers. Each requires tailored prevention tools. For example, stopping APP fraud may involve behavioural flagging, while account takeovers demand multi-factor authentication and anomaly detection.
Compliance is now a non-negotiable requirement. AML and counter-terrorism laws require platforms to know who they’re onboarding and how funds are used. Failures can result in financial penalties, regulatory scrutiny, or licence revocations. The UK’s Mandatory Reimbursement Bill has added another layer of urgency, mandating reimbursement for APP fraud victims and sharing the burden between sending and receiving payment providers.
Trust is also at stake. Repeated fraud incidents can prompt banks to block payment platforms entirely. If your platform becomes a known fraud target, partners may opt out, legitimate customers might churn, and regulators could clamp down.
Document fraud is a central enabler of payment crime. From Photoshop forgeries to deepfaked ID selfies, fraudsters use a wide array of tactics to bypass onboarding checks. Payment platforms must be able to identify altered, forged, or synthetic documents in both KYC and KYB processes. AI can help flag fake documents with impressive accuracy, under 1% false positive rates.
So how can fraud be prevented effectively? It comes down to three core components: secure onboarding, AI-powered transaction monitoring, and a layered defence that connects both.
Secure onboarding starts with enhanced KYC and KYB, using AI to detect synthetic identities and forged documents. This improves fraud detection without creating friction for genuine users.
AI transaction monitoring offers context-aware insights, connecting signals across a user’s journey. This enables platforms to spot unusual behaviour, suspicious flows, or repeated tactics used across synthetic mule accounts or high-risk transactions.
A layered defence combines both approaches, allowing systems to learn and adapt. This fusion protects against fraud across every stage—onboarding, account activity, and transaction monitoring—without hindering legitimate growth.
The benefits are wide-ranging: faster fraud resolution, lower operational costs, reduced regulatory exposure, and improved customer experience. Fraud caught early is less damaging and cheaper to fix.
Take Payoneer, for example. The company implemented Resistant AI’s document fraud detection and reduced manual reviews by 90%, catching synthetic identities and serial document fraud before accounts went live.
Fraud has evolved into a scalable, tech-driven enterprise. Criminals operate like start-ups—iterative, fast, and automated. Without the right tools, platforms are constantly behind.
AI is no longer optional—it’s the only viable way to match fraud’s scale and speed. It enables platforms to leverage KYC, KYB, and behavioural data to create a unified defence. This evolution in fraud requires an equally fast evolution in prevention.
Copyright © 2025 RegTech Analyst
Copyright © 2018 RegTech Analyst
