Companies in Central Asia are significantly outpacing the global average in adopting artificial intelligence in business. According to a KPMG study, in the region AI is much more often brought to industrial use and applied across several business processes at once. This was reported by Qazaqyia.kz citing Kursiv Media.
According to the study, 53% of companies in Central Asia and the Caucasus already use AI at industrial scale and achieve measurable results from multiple use cases. The global average is 24%.
For the report, KPMG surveyed more than 2,500 technology executives in 27 countries. The study included 32 respondents from Central Asia and 40 from the Caucasus.
At the same time, experts warn that some companies in the region may have interpreted the concept of industrial adoption too broadly.
"We assume that in our region respondents included in their answers such fragmented cases as text and voice bots," said Konstantin Aushev, partner and head of technology practice at KPMG Caucasus and Central Asia.
Despite this, Central Asia already has examples of large-scale AI use.
In Kazakhstan, one of the most notable examples is Kazakhtelecom. Since mid-2025, the company has been implementing AI in contact centers, customer service, procurement, HR, finance, document management, and occupational safety.
The company operates a hybrid model: a central team is responsible for platforms, security, and complex products, while employees can assemble simpler solutions themselves.
"For smaller tasks, we use an agent creation platform," said Yerzhan Assanov, Director of Artificial Intelligence at Kazakhtelecom.
Employees can describe a task without programming skills, upload work materials, and create their own digital assistant.
Business & Technology Services, part of Eurasian Resources Group, is also actively implementing AI. The company uses computer vision to monitor equipment at mining sites and AI to analyze geological data.
Experts attribute the rapid growth partly to the region's companies' propensity for experimentation.
According to KPMG, 44% of companies in Central Asia and the Caucasus use a proactive and experimental approach to AI. Globally, only 21% of companies do so.
Companies regularly launch pilot projects and quickly test whether they can be scaled to real business processes.
In BTS, for example, about 80% of projects based on classical AI — machine learning and computer vision — reach industrial implementation. For generative AI, this figure is currently about 10%.
At Kazakhtelecom, new AI solutions are tested almost every month.
"We launch experiments earlier and test them faster on real tasks," Assanov noted.
Another advantage of Central Asia, experts say, is the high level of business digitalization. Many large companies already have the necessary infrastructure, making it easier to connect AI to existing systems and data.
Despite high adoption rates, companies in the region lag significantly behind the global level in forming a unified AI strategy.
Only 34% of companies in Central Asia and the Caucasus said they have a full-fledged AI strategy linked to business goals. Globally, this figure reaches 76%.
The gap is even more noticeable in governance: a unified system for controlling and regulating AI exists in only 25% of companies in the region, compared to 69% globally.
According to Aushev, this is one of the main paradoxes of the market: companies actively launch AI solutions but are much less likely to define rules for their use and ways to evaluate effectiveness in advance.
Experts believe the next stage for Central Asian businesses will be a transition from numerous experiments to systematic implementation — with clear rules, cost assessment, and payback calculation.
Such a process is already beginning at the state level. In particular, Kazakhstan is gradually introducing more formalized rules in the field of artificial intelligence and the digital economy.