Resistant AI has raised $25 million for its financial crime/fraud prevention-focused artificial intelligence (AI) models.
The Prague-based company’s Series B round will let Resistant expand its document fraud detection and transaction monitoring products into new territories and partnerships, and further develop its threat intelligence capabilities, according to a Monday (Oct. 13) news release.
“The new funding comes as the anti-fraud and regtech market is being transformed by fully-native or bolted-on agentic solutions that replace static workflows with cheaper, smarter adaptive ones,” the company said in the release.
However, Resistant added, these large language model (LLM)-based agents are not designed to carry out the quantitative risk analysis required to combat fraud and financial crime (fincrime). They also suffer from high systemic hallucination rates of 10-30%, and have shown to be difficult to protect from “adversarial manipulation.”
Resistant says it works to “protect and empower” these AI agents and their “overwhelmed fraud, risk and compliance teams” with machine learning models that identify fraud in documents, transactions and behaviors.
“The financial crime landscape has fundamentally changed with the deployment of LLMs and AI agents in risk prevention settings, and the weaponization of generative AI by fraudsters,” Resistant AI founder and CEO Martin Rehak said in the announcement.
“Our fraud and fincrime models offer any institution the tools to empower both their human and agentic co-pilots to combat these AI-powered threats at scale.”
The company last raised funds in a 2023 Series A round, taking in $11 million. The latest round was led by DTCP, plus existing investors such as Experian, GV and Notion Capital.
PYMNTS looked at the challenges AI presents to financial institutions when it comes to fraud protection in a recent conversation with Jenna Kaye-Kauderer, managing vice president and head of AirKey at Capital One.
“With gen AI, there are new threats and new fraud vectors we hadn’t seen historically. And we’re seeing those today,” she said.
While fraud is not new, that report said, generative AI has changed its tempo, with criminals now able to build and scale convincing attack campaigns in minutes. That means the old system deploying a defense and assuming it works is no longer an option.
The discipline of authentication itself needs to evolve, “not as a product but as a continuous, adaptive practice,” PYMNTS wrote.
“One of the most important things for financial institutions and banks is first building flexible authentication systems that allow us to react dynamically as these new fraud vectors appear,” Kaye-Kauderer told PYMNTS. “And second, having a really broad set of authentication tools, a really big toolkit.”
Source: https://www.pymnts.com/