THOMAS the Fraubot

University Of Santo Tomas | Philippines

Kyle Edward Escosia | Ezekiel David | Ian Louise Sentina | John Cedric Custodio

Technology has significantly improved over the past century. It has allowed users to increase level of convenience when it comes to business transactions. Most people can easily transact online with ecommerce and even on physical store purchases. This was made possible with the use of electronic money and credit cards. Alongside with the rise of the technology, risks have also increased. Fraudulent transactions have been considered to be one of the most significant risks with the use of technology for purchases. Fraud has been a consistent pain for the users and corporate businesses alike, for the past years. This causes massive loss of money for both consumers and businesses. The most common activities involving fraud include, but not limited to the following: websites posing to be legitimate stores purchases requiring continuous subscription fees and premium payments without the knowledge of buyer valid purchases done from stolen credit card online transfer of electronic money from existing bank accounts valid transactions to aid money laundering In order to mitigate the risks and incidents of such scenarios, banks have been doing, mostly, manual, verification to the customers. Usually, business thresholds are set in terms of customer’s buying behavior or amount of purchase, then banks do their verification over the phone by calling the customer. One limitation of this however is when verifiers are already outside of working business hours which sometimes gives opportunity to fraudsters to commit their deed. Also, customers experience the difficulty of disputing fraud transactions if these were not reported or addressed immediately. It usually takes weeks and even more than a month to prove and reverse fraud transactions. Given the strong need to prevent fraudulent activities, the team has conceptualized to create a chatbot that has a capability to do educated and well-calculated verification with the user. The user will have little to no means to notice that this is a bot that they are chatting to. Such automated verification will prove to have very significant benefits to both users and banks. Details are further discussed in detail on the description of the idea.