Artificial intelligence and architecture of the global financial system

For 15 years The block chain -Blockchain-, interpellates the traditional centralized financial systemsupporting the development of the Universe Defi (decentralized finances) where diverse products such as bitcoin, stable cryptocurrencies, disintermedized platforms or programmable money navigate, among others.

By his part of the Quantum computing, in a state still embryonic for the public, promises to drastically change the concepts of efficiency and safetyto the point that the BIS (Bank of International Payments) has been warning of the central and commercial banks the imperative need to migrate their security platforms to postcussical cryptography.

The third technology, and in a currently relevant clear manner, It is the artificial intelligence that incorporates a wide variety of approaches and tools from different scientific fields With the purpose of creating systems endowed with intelligence, expressed in its various aspects are logical, emotional, verbal, artistic or social, among others.

The great AI toolbox and the financial system

The field of IA has evolved in recent decades thanks to a broad spectrum of developments and instruments Such as: Natural language processing -NLP-, broad language models -llm-, computer vision models, applied to robotics, internet of things -iot-, embedded, machine learning (ml), neural networks, deep learning, learning by reinforcement, chatbots, intelligent agents, knowledge representation, and others.

These developments do not behave as stagnant compartments. On the contrary, It is common to find all kinds of combinations of communicating techniques and vessels Among them.

Simplifying the purpose of facilitating the analysis referring to impact on the financial system, we can divide the AI ​​into two large branches, taking into account the main objective of its operation and the tasks that are performed: the generative the AI ​​on the one hand and the predictive one on the other.

In a very schematic classificationthe predictive AI analyzes massive amounts of past data to predict future dataand their applications to the financial world range from making credit, detect real -time fraud, or analyze investment risks based on historical information.

For its part The generative AI can produce original content (text, images, audio, video, code, etc.) from data and patterns learnedand may be used preponderantly to make financial reports, attend consultations by chat (chatbots) in a natural and contextual way, create or simulate possible economic or financial scenarios in the future or create educational or financial orientation content according to the user’s profile.

The future of AI in the financial system has evolved since the use of automatic learning (ML), through the development of AI generative and more recently incorporating the use of AI agents, -programs that can execute tasks autonomously- widely used by financial entities In recent years.

In terms of its applicability, the BIS published a report in June 2024 emphasizing that IA has improved the ability of the financial system to process information, analyze data, identify patterns and perform predictions, highlighting that automatic learning and deep learning models are widely used in assets assessment, credit qualification and risk analysis. Although Genai is incipient, the financial system is already adopting it to optimize administrative and business processes, customer service and regulatory compliance.

Algorithms, asymmetries and cognitive sovereignty

Much of AI rests on a basic and essential concept: The algorithms. Strictly an algorithm is a set of steps, logical, methodologically ordered, which when executed solve a problem or perform a task. From a practical point of view, an algorithm is the logic in the abstract of a program. It is the fuel that makes an application of the work.

These algorithms have managed to penetrate virtually all work or personal activities that we develop daily through the various devices with which we interact, be cell phones, computers, televisions.

Browsers such as Google Chrome, social networks such as Instagram, Facebook, X (formerly Twiter) or Tiktok; messaging services such as WhatsApp; Information and Communication Technology Companies such as Apple; E-commerce platforms as Amazon or MercadoLibre; streamings services such as Netflix are some close examples of how These algorithms affect our daily life using ia without us becoming fully aware of the degree of increasing influence that affects our behaviors and decisions.

In addition, and as regards the main actors, all these global use platforms that they use Ia are dominated by a very small number of large technology companies, “bigtechs” whose main asset is the gigantic database They build thanks to the information provided by their hundreds of millions of customers.

His Great capacity for innovation, a more lax and ethereal regulation than applied to its traditional competitors of the financial system, and recently the intensive use of AI They put the Bigtechs, in an advantageous competitive situation compared to the banking universe in what makes the provision of financial services.

With a multipolar geopolitical backdrop, North American and Chinese Bigtechs have launched into the conquest of this technology which encompasses a universe of multiple dimensions that interact with each other and long exceed financial guidelines and behaviors and behaviors and behaviors and behaviors.

The use of The gap or asymmetry that exists among those large companies that can process and use in advance That information before foot users.

In recent years, The growing concentration of technological power in the hands of a few companies, together with the massive use of personal information and practices such as algorithmic manipulation, It has generated concern in various sectors.

In response, a global movement of intellectuals and scientists from different disciplines has emerged that are elaborating the concept of cognitive sovereignty: A notion still under construction, which seeks to defend the informational autonomy of people, communities or states Faced with new forms of digital control.

Recent developments in Global Financial Services

The financial world In each of its various segments (banks, fintechs, financial and infrastructure services providers, and regulators) It has been one of the first to adopt the various AI tools.

Let’s look at specific cases in different global financial sectors:

• The largest western bank by asset volume, JP Morgan Chase, has developed an internal AI and automatic learning called Omniai That allows you to analyze and process large amounts of data, including transactions records, customer profiles, to identify anomalies, detect behavioral patterns and potentially fraudulent activities, automate processes and improve the experience of your customers. To calibrate the importance that this institution gives to AI, its headquarters Jamie Dimon has said that “we are completely convinced that the consequences [de la inteligencia artificial] They will be extraordinary and possibly as transformative as some of the main technological inventions of the last hundred years: think of the printing press, the steam machine, electricity, computer science and the Internet ”
• In the world of large technology companies, Bigtech Tencent, with more than one billion users in China, uses Wechat Pay Artificial Intelligence on their payments platform For multiple objectives. Among them, the advanced biometric authentication stands out, which has recently incorporated technologies such as the recognition of the palm of the hand. In addition, it uses AI to quickly evaluate loans to informal sectors without credit history, which significantly enhances financial inclusion.
• From its conglomerate – almost private – of companies, The controversial Elon Musk also participates in the race to achieve the financial “global supremacy”. To do this, he is integrating his artificial intelligence company, X AI, with its financial services platform X Moneya digital payment system based on the social network X (formerly Twitter). Musk seeks to transform this platform into a superplication capable of competing with the main bigtechs.
• As a particular case within the crypto ecosystem, on May 5, the general manager of the USDT issuing company –The main stable cryptocurrency linked to the dollar – announced the creation of Tether AI, a decentralized and open source platform that seeks to integrate artificial intelligence capabilities directly into personal devices. This platform will allow transactions in USDT and Bitcoin without depending on centralized or APIS servers (programming interfaces).
• Within the universe of large global financial services, Both Visa and Mastercard announced initiatives last month that incorporate artificial intelligence into their service offer. These innovations allow to integrate the process of making decision -making of their customers, channeling the payments gradually through autonomous software agents capable of understanding and executing user decisions.

In this sense, Visa presented its “Intelligent Commerce” platform, which allows the agents to token and process safe payments in a global network. For its part, Mastercard launched “Agent Pay”, which allows consumers to establish spending limits and permits for these agents, using tokenized credentials.

• As an example of global payment platformsPayPal recently launched its “agent toolkit”, a solution that allows the application to understand the client’s context and offer quick and personalized responses according to their situation. This tool enables any artificial intelligence agent to make payments, track shipments, issue invoices and solve disputes without the user having to leave the chat where he is interacting.

Risks and potentialities

Obviously, The advance of all these tools – to which the increasingly popular autonomous agents of AI are added in financial entities that decide and execute operations independently, based on their previous training– It raises huge risks, potentialities and complex challenges to address.

Two documents published this year by the bis referring to the role of central banks before these new Technologies urge regulators to adapt their structures as soon as possible and the formation of their human capital, warning the various risks derived from the use of AI. These risks that cover operational, information security, privacy and cybersecurity issues; Risks of Information and Communication Technologies (ICT); risks of third parties, such as external dependence; and risks inherent to AI models such as biases and the so -called “hallucinations” that occur when AI invents things that are not true.

While artificial intelligence expands accelerated and disorderly in multiple cases of use, And regulators try to arm themselves with adequate tools to face these complex innovationsthe various actors of the Argentine Financial System – Bancos, Fintechs, markets, service providers and financial infrastructure – navigate these waters trying to articulate a clear comprehensive strategy, trying to stay afloat in an increasingly efficient, technologized, concentrated market.

Carlos Weitz- former president of CNV and Technological Finance Professor. University of Buenos Aires.

Daniel Díaz – Postgraduate Professor Strategic Management of Computer Technology (GETI). Rosario National University.