AI will speed up processes in corporate bubbles
Artificial intelligence-based solutions could significantly speed up processes in the financial sector - especially those related to transactions in the corporate bubble. "On the regulatory side, however, there will be a lot to be said for the European Union and the Artificial Intelligence Act that is being developed, which could prohibit certain solutions and point the way forward for this innovative industry as a whole," says Jan Gąsiorowski, an associate in the Warsaw office of Wolf Theiss.
The AI Act, is a recommendation adopted this year for a future directive regulating the use of artificial intelligence-based solutions. It is still to be an area of negotiations at the EU forum and will come into force around 2026. However, it can already be expected that it will contain a number of restrictions, such as interference with privacy, among others, or prohibit solutions that could be considered disinformation.
Artificial intelligence, is an opportunity, for development and innovation, but at the same time a number of concerns related to it. For example, in corporate banking, it can be successfully used, for example, in processes related to the preparation of documents or the processing and analysis of key information, saving time and resources both on the part of the parties to the transaction and advisors, says attorney Jan Gąsiorowski, of the law firm Wolf Theiss.
The European Parliament began work on AI-related legal solutions about two years ago. Today, technological progress, has accelerated significantly, and the use of tools such as ChatGPT or Midjourney has not only increased the popularity of such technologies, but also aroused concerns about unethical or even illegal solutions.
The development of AI-based tools in corporate banking is currently at a fairly early stage, although I assume that a breakthrough could come unexpectedly and very quickly. Especially since we have quite a lot of data and repetitive, standardized procedures that can be automated. Personally, I think three areas are particularly interesting when it comes to the use of AI - these are solutions to support the due diligence process, issues related to the preparation and negotiation of contracts, and verification of documents provided by borrowers as part of conditions precedent to the disbursement of a loan," adds attorney Jan Gąsiorowski.
The report "Banking in 2035: a global survey of the banking sector," prepared by Economist Impact, shows that for 55% of respondents, the digitization of financial services and the introduction of new technologies are among the most important trends in the banking industry. In turn, the McKinsey report "The state of AI in 2022" reports that already more than 52% of businesses worldwide are allocating 5% or more of their IT budgets, to the development of artificial intelligence, while 63% of companies plan to increase their spending in this area in the coming years.
If we now look at the development of technological tools, including those based on artificial intelligence, in online banking, we can expect the expansion of these and similar solutions into other sectors of banking services - including, of course, corporate banking. Problems may arise, however, in sensitive areas that require specialized, expert advice such as complex tax issues. On the other hand, we already have experiments related to the use of AI in valuing assets, or interpreting financial data. I think the next step will be the use of AI to perform analysis and verification of all kinds of documents provided by the borrower as part of the conditions precedent to loan disbursement. This kind of solution would be particularly useful for analyzing leases that are provided to the bank in commercial real estate financing transactions. However, the question remains as to who will be responsible for any errors and oversights of AI," stresses attorney Jan Gąsiorowski.
According to a European survey conducted by IPSOS in 2020. - civil liability is one of the three biggest barriers to using AI-based technologies.
We look at how big a problem this is for the medical industry, where liability for potential errors is key. We have also had reports related to autonomous cars, which can prove dangerous to other road users. In the Netherlands, on the other hand, a scandal erupted in 2019 over benefit checks. As it turned out, as a result of falsified baseline data, the authority used algorithms to apply sanctions for extortionate benefits only to a very specific ethnic group of people. In corporate banking, we can also easily imagine a malfunctioning algorithm if the data used to create it turns out to be wrong or outdated in a different or dynamically changing economic situation. Hence, there are natural concerns about the development of AI in many sectors of the economy, which, however, should not impede the development of these technologies, notes attorney Jan Gąsiorowski.
As envisioned in the current draft of the Artificial Intelligence Act, in the EU we are to distinguish four risk thresholds for the application of AI - ranging from unacceptable to minimal. How a tool or area is classified will also determine the responsibility for the solution used. The highest level involving, among other things, the risk of manipulation is to be banned in the EU, while responsibility for the operation of high-risk systems will fall on suppliers. Significantly for banking services, the EU also considers scoring, which is the classification of people based on their behavior, socioeconomic status or personal characteristics, to be extremely sensitive.