Mastercard, which is one of the largest players in the payment industry, announced the creation of its artificial intelligence model.
The AI configuration developed by the mentioned firm is designed to provide thousands of financial institutions with the ability to detect fraudulent transactions in their functional networks. Also, the generative artificial intelligence model created by the giant of the payment industry can eradicate the specified activity.
The company told reporters that its development was called Decision Intelligence Pro. Banks, using a generative artificial intelligence model, will determine transactions in real-time in terms of their compliance with such a parameter as the security of the procedure. These actions will be performed in real-time. AI will help banks make the right decision when determining the legality of financial operations.
Ajay Bhalla, president of the Mastercard cybersecurity and intelligence business unit, says that the new artificial intelligence solution is a patented recurrent neural network. He noted that the generative AI model is an independent development of the company’s team of specialists, whose activities are related to ensuring security in cyberspace and combating fraud.
Ajay Bhalla, during a conversation with media representatives, said that the giant of the payment industry uses transformer models that help apply the capabilities of generative artificial intelligence. According to him, the new development is embedded in the functional space of Mastercard, where there are all kinds of data from the brand ecosystem. He stated that from the very nature of the business that the e-commerce giant is engaged in, the company sees all the information about transactions that come from the ecosystem.
According to Ajay Bhalla, in some cases, Mastercard relies on open-source code as needed. At the same time, he noted that most of the technology is created by the payment industry giant’s own efforts.
Mastercard’s algorithm is trained on data from about 125 billion transactions that pass through the company’s card network annually. Using arrays of information helps artificial intelligence understand the relationship between sellers, rather than words, as in large language models such for example, GPT-4 from OpenAI and Google Gemini. The giant of the payment industry claims that its generative AI model predicts where fraudulent transactions occur.
The Mastercard algorithm, instead of arrays of text information, uses the history of the cardholder’s visit to the store as a hint to determine whether the business involved in the financial operation is the place where the consumer is most likely to go. This algorithm generates pathways across the e-commerce giant’s network to get answers in the form of a score. To some extent, the mentioned mechanism of action is similar to the principle of operation of a heat-sensing radar.
A higher score will match the pattern of normal behavior expected from the cardholder. A low score gets a negative result during what can be described as a compliance procedure.
Mastercard reports that the algorithm performs all the necessary actions in just 50 milliseconds.
Ajay Bhalla says that the new technology of making decisions about transactions from the giant of the payment industry can help financial institutions increase fraud detection rates by an average of 20%. According to him, in some cases, this solution increased the corresponding figures by an impressive 300%.
Mastercard has invested more than $7 billion in cybersecurity and artificial intelligence technologies over the past five years. In this case, it also implies several acquisitions made by the giant of the payment industry, including the purchase of the Swedish firm Baffin Bay Networks in March last year.
The problem of cybersecurity is becoming more urgent and sensitive as the number of individual actions and large-scale processes performed by people in the virtual space increases. In this case, risks arise not only when implementing certain procedures in a digital environment, but also during the operation of personal devices. Cybercriminals can gain access to gadgets whose owners are unaware of outside interference. In this case, awareness of the methods and signs of illegal access to the device can become a counteraction tool. The relevant data can be obtained, for example, on the Internet through the search query How do I know if my phone camera is hacked? In such cases, cybercriminals not only gain access to confidential information but also carry out surveillance of their personal lives.
Returning to the topic of artificial intelligence in the sphere of financial services, it should be noted that Mastercard’s competitor, which is Visa, is also investing in AI, including a $100 million venture fund for startups in the area of generative machine intelligence. This fund was established in October last year.
Mastercard expects its AI algorithm to allow banks to save up to 20% by reducing most of the costs associated with evaluating illegal transactions. Ajay Bhalla says that the true potential of the company’s technology lies in the ability to identify fraudulent schemes and trends to predict future types of relevant crimes that are still unknown in the payment ecosystem. According to him, the advantage of the Mastercard ecosystem is that the brand sees data on financial operations from all customers around the world. In this case, the company can form a full and up-to-date understanding of the specifics of the procedures and detect suspicious activity based on this information.
Firms that operate in the sphere of payments and digital banking are increasingly claiming that artificial intelligence will seriously change their products. Last week, PayPal announced new products based on machine intelligence.
