Payment identity network iPiD has joined forces with Microsoft to builds an AI-powered name-matching tool for Verification of Payee (VoP).
Name matching has long been a cornerstone of payment security and regulatory compliance. Financial institutions are tasked with ensuring accuracy to prevent processing payments for sanctioned entities.
New European rules, set to take effect in October, add new VoP complexity, requiring financial firms to verify that the name provided by the payer matches the name registered to the recipient’s bank account in order to combat payment scams.
Payee's financial institutions will have to classify name matches onto three categories before a payment is processes: match; partial match; or no match.
Partial matches involving minor discrepancies — like spelling errors or alternate naming
conventions — that may still suggest the correct payee, present an issue for banks. While providing partial matches improves customer experience, it raises GDPR concerns when inadvertently revealing unintended payee information.
To address this, iPiD has tapped into the MicrosoftAzure OpenAI Service to build a name matching algorithm specifically designed to meet the complex requirements of VoP. The firm claims it balances precision and compliance with speed and efficiency, reducing false negatives while adhering to strict data privacy requirements.
Ivan Chong, CTO, iPiD, says: "Names are more than just letters; they carry identity and meaning. With AI, we can bridge the gap between technical precision and contextual nuance.
"Thanks to our collaboration with Microsoft, we’ve created a solution that combines the linguistic depth of large language models with essential requirements like speed, resource efficiency, explainability, and adaptability."