Concept: US-based tech startup FiVerity has launched a new machine learning solution, Collaborative AI platform, to detect and prevent cybercrime. It helps banks, credit unions, credit cards and other financial service providers in combating synthetic identity fraud (SIF). The platform identifies fraudsters using a pattern matching technique without needing consumer personally identifiable information (PII).
Nature of Disruption: FiVerity uses a collaborative AI platform that collects the customer account data submitted by the financial institutions to verify the applicants. The solution then provides scores for each application considering the level of risk instantly. The institutions can analyze and reject the high-risk applications and review the ones that seem suspicious. The platform leverages legacy rules-based methodologies for reviewing and deciding whether to reject the applications. It uses machine-learning to learn from the patterns from the customer data sets to build their prevention and detection capabilities. The platform has detected more than 35% of the undetected SIF for financial institutions utilizing machine learning solutions. The AI platform can perform customer onboarding, application processing and credit verification processes, helping banks and other institutions to streamline and scale their business operations.
Go deeper with GlobalData
Outlook: The major threat faced by banks and financial institutions is SIF, which causes losses in the industry. Identifying criminals who create fake identities and produce false information to avail loans from the institutes is a time-consuming process because they use automation tools to create identities and banks could not identify such cases. FiVerity’s AI platform acts as a central resource that helps banks and credit unions to identify fraudsters by analyzing the customer account data. The startup has plans to provide the solution to other industries and government agencies after the completion of the initial pilot phase.
This article was originally published in Verdict.co.uk
