Around 60% of companies in Kazakhstan could integrate AI into core processes

Over the next 2–3 years, AI agents could become part of the everyday IT infrastructure of large and medium-sized businesses. IBA Group estimates that around 60% of organisations in Kazakhstan will use them to redesign core processes.

Over the past six months, the number of requests to implement AI assistants has increased by approximately 30%. Most of the demand comes from large businesses seeking to automate routine operations and speed up decision-making.

Corporate AI systems can analyse internal documents, review contracts, prepare financial reports, and identify anomalies in business processes. They are also used to process customer enquiries and search for information in corporate knowledge bases.

Over the next 3–5 years, AI agents could become as common a part of large companies’ infrastructure as ERP systems. Individual agents will handle procurement, supplier analysis, reporting, regulatory compliance, and employee support.

McKinsey estimates that generative AI and other technologies could theoretically change how tasks accounting for 60–70% of working time are performed. This refers to automating or supporting specific operations rather than replacing the same proportion of employees.

The financial sector remains one of the most active users of AI. The technology is used for credit scoring, document verification, fraud detection, risk management, and customer service.

According to the National Bank of Kazakhstan, 36% of surveyed financial organisations already use AI, while 56% are considering adopting it within the next year. Only 2% have reached full-scale implementation.

The study covered 232 financial-sector organisations in Kazakhstan, Kyrgyzstan, and Tajikistan. These figures therefore reflect the Central Asian financial market rather than all companies in Kazakhstan.

In retail, AI can forecast demand, manage product ranges, personalise offers, and reduce excess inventory.

Manufacturers can use intelligent models to predict equipment failures, reduce downtime, and monitor product quality. In logistics, the technology can help plan routes, allocate vehicles, and shorten delivery times.

The main barriers to AI adoption remain a lack of high-quality data, outdated IT architecture, and unprepared business processes. Companies also need to train employees to work with AI assistants and determine which decisions should remain under human control.

Businesses are gradually moving away from isolated experiments towards processes with a clear economic impact. Before scaling, companies will need to prepare their data, define performance metrics, and integrate AI solutions into their existing operating models.