EAEU central banks have begun discussing AI regulation in financial markets
Country and market
For Central Asia, particularly the banking sectors of Kazakhstan and Kyrgyzstan, the issue of algorithm regulation is long overdue. Local banks and fintech companies are already actively using machine learning. Neural networks approve loans, analyze transactions for fraud, and conduct biometric customer identification.
Until now, the implementation of such solutions has occurred in a regulatory vacuum or relied on general information security rules. The discussion at the Advisory Council level shows that central banks are ready to transition from passive observation to forming a specialized regulatory framework.
Why it matters
The inclusion of technologies in the central banks’ agenda means that regulators are preparing a real foundation for standardizing regtech and suptech solutions so that bank algorithms do not turn into an opaque black box.
When a bank denies a customer a loan or blocks a transfer due to a neural network’s decision, the regulator needs to understand the algorithm’s logic. Without common standards for risk analytics and KYC/AML procedures, the financial systems of different countries will develop in isolation. Creating clear rules of the game will allow fintech companies to scale their products more easily within the region.
What’s next
The current discussion is advisory in nature. In the near term, central banks will have to develop a unified conceptual framework and determine exactly which AI models require strict supervision and which can develop freely. It is likely that credit scoring algorithms and anti-money laundering systems will be the first to fall under regulation.