Startups Use AI & Human Insight for Fintech Fraud Detection
Hasan Hakim Brown used fake identities to defraud institutions and the Small Business Administration of more than $20 million. He combined stolen Social Security information with fabricated names to elude fraud-detection systems and utilized multiple virtual desktops. In contrast, the San Francisco-based startup SentiLink aimed to combat synthetic ID forgery. The rise of a fintech startup, established by Maxwell Blumenfeld and Naftali Harris, happened with an estimated $25 million in revenue and a $430 million valuation in 2022. They offer AI-powered fraud detection solutions but emphasize the necessity of human supervision in recognizing patterns and developing fraud techniques. SentiLink has over 300 customers, including prominent banks, fintech companies, and credit unions.
Harris and Blumenfeld became interested in synthetic fraud while working as data scientists for Affirm. Blumenfeld uncovered multiple borrowers with the same name and birthday but specific Social Security information, with excellent credit histories and scores. They realized that fraudsters were extracting Social Security information from less probable credit applicants, merging them with fake identities and genuine addresses, and establishing credit records to qualify for big loans they did not intend to repay.
SentiLink utilizes ML (machine learning) as a critical aspect of its business to combat synthetic ID fraud. It employs 78 individuals, eight devoted to manually evaluating fraud attempts and the others to identifying novel trends or potential false declines by the AI algorithm. The founders emphasize that their artificial intelligence model aims to imitate human behavior rather than automatically detect fintech fraud. The company’s success stems from its inventiveness, speed, and capacity for rapid expansion.
To keep up with the latest developments in the world of fintech and for more fintech news, please visit https://worldfintechnews.com/.