rewrite this title Feedzai Launches RiskFM to Enhance Financial Crime Detection – Finovate

Financial crime and fraud prevention specialist Feedzai unveiled its RiskFM (Risk Foundational Model) solution this week.

RiskFM covers a broad range of financial data to provide risk decisioning across fraud detection, anti-money laundering, and other financial crime.

Headquartered in New York and founded in 2008, Feedzai made its Finovate debut at FinovateEurope 2014 in London.

Financial crime prevention innovator Feedzai introduced its RiskFM (Risk Foundational Model) solution this week. The new offering leverages a Tabular Foundation Model that is purpose-built for financial data and risk decisioning, changing the way that financial crime is detected and prevented.

Spanning across fraud detection, anti-money laundering (AML), and other financial crime-related risk decisions, RiskFM is trained on a broad, deep, global dataset covering onboarding, digital activity, payments, fund transfers, and AML workflows to enable institutions to identify, prevent, and adapt to financial crime quickly and accurately.

The solution is designed to handle some of the special challenges of dealing with transactional data. In their statement announcing the new offering, Feedzai compared this challenge with large language models (LLMs) and their ability to deal with domains such as language, audio, and video. These domains, the company noted, all have finite grammar and a certain linear causality. By contrast, financial transactions are far less predictive, in large part because the consumer behavior behind these transactions, from payment types to fraud modalities, can and does change—frequently.

“Next transactions are far less predictable than the next word in a sentence,” Feedzai Chief Science Officer Pedro Bizarro said. “Consumer spending habits, payment types, and fraud modes change continuously. More importantly, financial risk is an adversarial domain; fraudsters actively adapt to evade detection in real time.”

The ability to operate across multiple institutions and geographies at the same time is one key feature of RiskFM, and when used to power a customized model for a single customer, RiskFM matches the performance of high-tuned, supervised models while avoiding time-consuming, manual feature engineering. RiskFM outperformed traditional models based on Gradient Boosting and Deep Learning strategies, and is built for the full range of financial crime prevention, from mule account detection to anti-money laundering. The company refers to the technology as the “foundational AI layer for financial risk,” ensuring institutions have an intelligent, scalable solution that grows as they do.

“RiskFM proves our multi-year investment in foundation models is paying off,” Feedzai Chief Product Officer Pedro Barata said. “We’re not just part of the conversation; we’re defining how it applies to the complexities of global financial crime prevention.”

Feedzai made its Finovate debut at FinovateEurope 2014. Headquartered in New York and founded in 2008, Feedzai today offers an AI-native financial crime prevention platform that helps banks, payment networks, acquirers, and other financial services providers detect and prevent financial crime, fraud, and money laundering in real time. The company’s platform serves more than one billion consumers, processes 90 billion events, and secures $9 trillion in payment volume annually.

In the wake of its RiskFM announcement, the company since reported that it has been named to Fast Company’s World’s Most Innovative Companies 2026 roster. “We at Feedzai are honored by this prestigious recognition of our innovation and research in trusted AI to build a world of safer money,” Feedzai Co-Founder and CEO Nuno Sebastiao said.

Photo by Tima Miroshnichenko

Views: 20