Fincom’s Anti-Discriminating service enhances any AML system with an ability to Establish a Reasonable Linguistic and Phonetics Justification when matching a searched name to a name that appears in Sanction & PEP lists. This includes handling unstructured and non-English names and changes in the order of the name parts. (e.g. placing the last name first, middle name last, etc.), as an example: Jose-Luis-Gonzales, Mohammad-Abu-Bakri-El-Baghdadi, etc. Fincom provides the only technology that screens the entire name (first-middle-last or any other structure) with unprecedented level of accuracy, guaranteeing ‘accurate name match’ and therefore preventing AML-Discrimination.
Fincom’s ADM can be installed as a ‘suppression engine’ with any existing AML system, filtering out results that cannot show justifiable phonetic proximity, helping, at the same time, to reduce the False Positives rates by approximately 30%.
Most of AML solutions on the market are technologies that lead to AML-Discrimination. It is an “inherent ethnical, religious, or other kinds of Bias” based on or associated with the name. Financial institutions are caught between the pressure to comply with AML regulations and the threat of extremely high fines on the one hand, and, on the other hand, customer complaints and penalties for discrimination, not to mention public pressure for social responsibility. The problem is that every Mohammad, Vladimir, or Jose-Lopez are alerted as possible terrorists, gangsters, or drug lords – yet these names are extremely common and do not in any way or manner indicate such criminal association. In other words, taking a name that is constructed from first-middle-last (and even more complex name structures) and basing a match on a partial section of a name cannot really produce a sufficient identifier. Trying to resolve this problem by using AI/Rule-based technology aggravates the problem.
Since most of these commonly used falsely alerted names are affiliated with some religion or nationality, it creates an ‘AML Bias’ that may naturally lead to legal claims against a bank for ‘customer Discrimination’. The existing AML systems do not have a tool that could create a solid a reasonable phonetic and linguistic justification for alerting or flagging those names.
Fincom’s ADM successfully resolves the challenges of AML- and AI-Discrimination by transforming names into mathematical codes based on the sound of a spelled name.