The Bank for International Settlements shared its opinion and vision of how financial institutions, primarily those of these organizations that have the status of state regulators, should respond to the active spread of artificial intelligence and interact with this advanced technology.

Central banks need to improve their grip regarding machine intelligence. This is the opinion of the experts of the Bank for International Settlements, which is based in Basel. In the context of the specified point of view, it was clarified that the financial regulators of the countries of the world should increase the level of control of artificial intelligence to assess its economic impact and use advanced technology for themselves.

On Tuesday, June 25, the Bank for International Settlements released a report that provides an assessment of machine intelligence in the context of various aspects of the functioning of the digital mind and scenarios of its impact on workflows and the final result of these processes. Experts of the mentioned financial institution are convinced that artificial intelligence will not replace human judgments. At the same time, they acknowledge the likely significant impact of advanced technology on productivity, investment, and consumption. According to them, the corresponding prospects for scaling artificial intelligence as a source of transformation or performance change mean policymakers need to study AI to subsequently use digital thinking systems for their own purposes.

The report of the Bank for International Settlements notes that the rapid and widespread introduction of machine intelligence implies that financial regulators should take measures as soon as possible to raise their game. In this context, it implies the need to improve the capabilities of central banks both as informed observers of the consequences of technological advances and as direct users of technology.

Reflecting on the topic of such an aspect of the impact of artificial intelligence on workflows as productivity, experts from the Bank for International Settlements noted that, for example, software developers using AI can more than double the number of projects that they coding within a week. Also, in their opinion, within the framework of the prospects for the near future, a scenario involving the growth of companies’ investments in information technology is likely to be implemented.

Moreover, experts from the Bank for International Settlements noted that for consumers, the expansion of the use of artificial intelligence means that algorithms will be developed that provide better choices in decision-making regarding spending-related actions. In this context, they also noted that the mentioned aspect of the new technological reality formed by machine intelligence will contribute to the growth of consumption volumes.

Cecilia Skingsley, head of the Bank for International Settlements Innovation Hub, pays attention to the significant potential associated with the ability of the artificial intelligence algorithm to process huge amounts of information. In this case, it is implied that the mentioned ability means the aptitude to solve tasks such as forecasting the dynamic of changes in the current situation. For central banks, the use of artificial intelligence algorithms for specified purposes forms the prospect of making more effective decisions.

Hyun Song Shin, an Economic Adviser and Head of Research at the Bank for International Settlements says that setting interest rates is likely to remain a task that people will solve since judgments are needed in this case.

Cecilia Skingsley, in the context of thinking about the use of artificial intelligence, drew attention to the fact that the AI algorithm cannot be responsible for its actions. According to her, the concepts of organizing processes in the space of vital activity of society and a similar methodology concerning the life of an individual provide for holding people accountable for decision-making. In this context, she mentioned the fact of the turnover of politicians and governors of central banks. Cecilia Skingsley does not see such a configuration of the future in which the issue of interest rates will be a task for solving by an artificial intelligence system.

The report from the Bank for International Settlements also notes that the impact of machine intelligence on inflation will depend on expectations of welfare gains and will differ in the short and long term. Experts at this financial institution believe that as long as consumers underestimate the future benefits of using artificial intelligence, new opportunities will be a factor constraining price growth. According to them, the corresponding tendency will be on the trajectory of movement in the opposite direction when people begin to fully assess the potential of machine intelligence.

The report of the Bank for International Settlements also notes that the development of artificial intelligence will generate new risks for central banks. In this case, cybersecurity is implied. For example, artificial intelligence can be used as a tool to commit cyber attacks. Also recently, AI has increasingly been used as a kind of disinformation generator. This does not mean that the information materials generated by artificial intelligence contain lies and incorrect mental representations of facts and phenomena of reality. In the context of the problem of disinformation, it is implied that AI can be used for deepfakes. Artificial intelligence, within the framework of a certain applying scenario, generates content with a high level of realism, but containing false or manipulative data. In this case, the activity is based on user commands. Deepfakes can mimic the voice and demeanor of certain people, forming their most realistic AI prototype in videos. Such content is used to implement fraudulent schemes or to spread narratives in the information space designed to strengthen in the public consciousness specific interpretations and explanations of the processes taking place in the plane of objective reality as absolutely true and containing the undeniable truth.

It is worth noting that against the background of the active spread of artificial intelligence and its deep integration into various processes, the issue of cybersecurity has become especially relevant not only at the level of operating of financial institutions or any organizations in other spheres of activity but also in the so-called household dimension. Scammers often use AI to deceive social media users. Artificial intelligence can also be applied by cybercriminals to communicate with victims on various digital platforms on behalf of their superiors, relatives, or friends. By imitating the voice of a real person, AI can make the interlocutor believe that all the information being reported is absolutely true. In this case, the personal awareness of users can become an important tool for countering fraud committed in the virtual space. For example, a query in an Internet search engine, such as how to know if my camera is hacked, will allow anyone to get information about signs of unauthorized access to the device. Digital literacy is what can be described as a tool for detecting the facts of cybercrime.

Experts from the Bank for International Settlements also note that when implementing a scenario in which most market participants use the same set of artificial intelligence algorithms from leading companies specializing in cloud computing, a threat to financial stability may form and even there is a possibility of unintended collusive behavior. It should also be added to these considerations that in the context of such a state of affairs, what can be described as the scalability of the destructive effect of potential failures arises. In a practical sense, the mentioned problem means that with certain problems in the functioning of the artificial intelligence service, which most market participants rely on, the negative consequences will be widespread and will affect a large number of consumers. There is also a risk that the company serving the majority of users may begin to abuse its position.

The use of artificial intelligence in the activities of financial institutions is a process that simultaneously generates opportunities and risks. The report of the Bank for International Settlements notes that the mentioned realities are particularly sensitive for areas of payments, insurance, lending, and asset management. The opportunity factor varies depending on improving the effectiveness of risk assessment and portfolio allocation to liquidity management. Potential challenges for banks in the context of scaling up the use of artificial intelligence are the impact of cyber attacks, the herd behavior of AI algorithms, and privacy concerns.

The report of the Bank for International Settlements notes that in the payment area, the abundance of data at the transaction level allows artificial intelligence models to overcome long-standing pain points. For example, the ability of AI to understand potential risks can increase the effectiveness of combating money laundering practices.

The report of the Bank for International Settlements also notes that in credit assessment and lending financial institutions have been using machine learning for many years, but artificial intelligence in this case generates additional opportunities. For example, AI can improve credit scoring by applying unstructured data. Artificial intelligence-based tools allow lenders to assess the creditworthiness of customers using arrays of alternative information. In this case, it means data such as transactions on consumer bank accounts or information about rental housing.

Separately, the report of the Bank for International Settlements notes that artificial intelligence is actively used in insurance, in particular for risk assessment and pricing. AI can also be a tool for automatic analysis of images and videos to assess the damage caused to property as a result of natural disasters.

In asset management, artificial intelligence models are used to predict returns. Also, in this case, AI is a tool for eliciting unobservable features of financial data (so-called asset embeddings. This scenario of using artificial intelligence in the financial industry is mentioned in a report by the Bank for International Settlements. It is noted that AI models provide a better understanding of the risk-return properties of portfolios.

To mitigate risks and harness the potential of machine intelligence, monetary authorities should share experiences, tools, and data. The report of the Bank for International Settlements calls for appropriate action. Experts from the financial institution say that central banks urgently need to cooperate to develop a community of practitioners. In this context, it is noted that against the background of rapid technological changes, the exchange of information on policy issues arising from the role of financial regulators as data producers, users, and disseminates is crucial. It is also underlined that collaboration helps to lower costs, as the community will foster the development of common standards. It is separately noted that central banks have experience of successful cooperation to overcome new challenges. The emergence of artificial intelligence has accelerated the need for collaboration in the data sector and the management of information arrays. This is the opinion of the authors of the report of the Bank for International Settlements.