Oscilar’s AI-powered ACH Fraud Detection identifies and prevents fraudulent transactions

Oscilar released its AI-powered ACH Fraud Detection product. The solution identifies and prevents fraudulent transactions with unparalleled speed and accuracy by leveraging advanced machine learning algorithms, generative AI techniques, and real-time data analysis and explainability.

This innovative solution is critical as the ACH Network – the backbone of the modern financial ecosystem – experiences unprecedented growth, with 8.2 billion payments handled in the first quarter of 2024 alone, including a staggering 47% increase in Same-Day ACH volume. ACH facilitates critical transactions such as payroll, bill payments, internet purchases, person-to-person transfers, and business-to-business settlements.

ACH credit fraud increased by 6% between 2021 and 2023, and more than half of organizations with revenue less than $1B were unable to recover funds lost from fraud attacks.

“Oscilar’s AI-powered ACH Fraud Detection solution has been a game-changer for our company,” said Maurice Harary, CEO of Fluz. “We were facing a significant challenge with first-party fraud, where bad actors set up accounts with neobanks intending to defraud us. They would deposit the funds and then withdraw them before the ACH pull. We were on the verge of having to shut down neobanks as a funding source. Oscilar’s advanced fraud detection model quickly identified these instances, enabling us to take swift action and prevent substantial losses.”

As the ACH Network experiences rapid growth and faster payments, the threat of ACH fraud has become more pressing than ever, with organized groups, data breaches, and sophisticated techniques exploiting vulnerabilities in the system. Fraudsters employ a wide range of tactics, including first-party fraud, account takeovers, stolen account details, scams, Business Email Compromise (BEC), vendor and payroll impersonation, money mules, and check kiting, which are increasingly complex and difficult to detect using traditional fraud prevention methods.

The FBI reported that BEC scams remain highly prevalent, resulting in $2.9 billion in losses in 2023, making it the second-costliest type of cybercrime.

“We founded Oscilar to make the internet safer and protect online transactions,” said Neha Narkhede, CEO of Oscilar. “Cracking down on ACH fraud is one of the biggest challenges for our fintech and financial institution customers, and ACH fraud costs companies and consumers billions each year. Our new ACH Fraud Detection product allows customers to prevent fraud in real-time while also helping them navigate the regulatory landscape with ease and confidence.”

Traditional fraud detection solutions have struggled to keep pace with the ever-evolving threat landscape. They rely on outdated rules engines or static Machine Learning models and manual processes that are slow, inefficient, and prone to errors. This has exposed financial institutions and businesses to significant financial losses and reputational damage.

As the ACH system has modernized, attackers have sharpened their tools and techniques, making it increasingly challenging to detect and prevent fraud. Oscilar’s ACH Fraud Detection solution addresses six key types of fraud:

• First-party fraud: Oscilar’s advanced first-party fraud detection model analyzes transactional data, customer history, and account status to identify instances where customers intentionally initiate ACH transactions with insufficient funds or falsely claim they didn’t authorize bill payments or transfers.
• Account takeover: Oscilar monitors device IDs and behavioral patterns to detect fraudulent ACH transactions initiated from compromised banking app accounts, protecting customers’ funds from unauthorized access.
• Stolen account details: Oscilar’s sophisticated Machine Learning algorithms identify unauthorized ACH transactions initiated from outside the banking app with stolen credentials, enabling financial institutions to detect and prevent fraudulent activities occurring off-app.
• Scams: Oscilar’s scam detection capabilities analyze transaction patterns, notes and recipient information to flag suspicious ACH transactions related to fraudulent schemes, safeguarding customers from falling victim to scams.
• Business Email Compromise (BEC): Oscilar’s AI-powered solution helps detect BEC, vendor impersonation, and payroll impersonation frauds by analyzing transactional data and customer behavior patterns. By identifying anomalous payment requests and suspicious transaction patterns, Oscilar helps prevent unauthorized payments from being “pushed” from a payer’s account to a fraudster’s account.
• Money mules: Oscilar leverages receiver reputation signals and account behavior monitoring to identify money mule accounts used to receive and transfer fraudulent funds, disrupting fraudulent networks and preventing further financial losses.
• ACH check kiting: Oscilar detects ACH check kiting schemes where criminals move money between accounts at different banks, taking advantage of the clearing house process to create the illusion of available funds while the money has already been withdrawn.

Oscilar’s holistic approach to ACH fraud detection includes real-time orchestration capabilities, advanced feature engineering for behavioral profiles, hybrid machine learning models, and AI co-pilots for Risk teams. These capabilities not only detect new attack vectors fast, but also help risk operations teams scale their investigations and reviews.

This comprehensive approach is well-suited for detecting and staying ahead of the ever-evolving nature of ACH fraud. It enables financial institutions to effectively combat fraudulent activities and protect the entire customer journey before money moves out.

Traditional fraud detection methods, such as identity matching, balance checks, two-factor authentication, document verification, and SSN validation, often result in low approval rates and require manual review, leading to increased friction for legitimate customers. Oscilar’s AI-powered solution goes beyond these conventional methods, analyzing bank account usage patterns and validating intent to identify and prevent fraudulent activities, such as the creation of accounts for the sole purpose of committing fraud.