Kazakhstan’s Agency for Regulation and Development of the Financial Market (ARDFM) plans to establish a regulatory framework governing the use of artificial intelligence (AI) in the country’s financial sector. The announcement was made by ARDFM Chair Madina Abylkassymova during the 13th Congress of Financiers of Kazakhstan.
According to the regulator, 75% of banks in Kazakhstan already use AI technologies in areas such as credit scoring, fraud detection, marketing, and customer service chatbots. However, adoption levels vary significantly: large banks have made substantial progress, while medium and smaller institutions are primarily conducting pilot projects.
Abylkassymova stressed that the widespread use of AI is imminent, necessitating the development of unified standards and regulatory guidelines.
“There is currently no dedicated document regulating artificial intelligence, but one will certainly be introduced,” she said. “The main challenges today involve the absence of uniform standards, inconsistent data quality, and a shortage of qualified specialists. At present, each financial institution independently sets objectives and configures its AI systems, but without common rules. We must therefore rely on existing regulatory practices and client-service standards.”
She added that regulatory measures should be proportionate to the level of risk associated with specific financial transactions. “We will conduct a detailed risk assessment. We do not believe every application of AI must be tightly regulated. Low-risk areas may require light oversight, whereas high-risk applications, particularly those with systemic implications, should face stricter controls.”
Abylkassymova identified data quality as a major obstacle to AI development. She noted that information is often fragmented and stored across various systems and formats, increasing the likelihood of errors in AI-driven decision-making. A proposed legislative initiative includes the creation of a unified database of anonymized data accessible to market participants, aimed at improving AI model accuracy.
Ongoing discussions are addressing whether access to government databases should be free or paid, and whether different pricing models should apply depending on the purpose of use.
“We place extensive requirements on financial institutions, including offering access to government services through their digital platforms,” Abylkassymova explained. “To fulfill these functions, banks need access to public databases. So, the question is: is it justified to charge for such access? That’s one scenario. It’s another matter entirely when access is sought for purely commercial gain.”
She concluded that the future AI framework will be developed in consultation with market stakeholders, with the goal of striking a balance between enabling innovation and safeguarding financial system stability.
Previously, The Times of Central Asia reported that Kazakhstan’s Ministry of Finance is piloting a digital platform that leverages AI and big data to help entrepreneurs identify the most profitable sales locations for their products.
