Targeted AI adoption could materially boost Kazakhstan's value added and employment: simulations attribute 16.8 percentage points of a projected 35.3 p.p. GVA increase (2025–2035) to AI, lift the GDP share of priority sectors by 6.3 p.p., and create roughly 1.3 million jobs, as task-creation effects outweigh automation.

The Impact of Artificial Intelligence on the Structural Transformation of Kazakhstan's Economy

Y. Varavin, M. V. Kozlova, M. U. Rakhimberdinova, Özay Özpençe · April 05, 2026 · The economy strategy and practice

openalex other low evidence 7/10 relevance DOI Source PDF

A structural simulation calibrated to Kazakhstan industry data projects that AI-driven adoption across priority sectors could contribute 16.8 percentage points to a 35.3 p.p. cumulative GVA gain by 2035, raise the priority-sector GDP share by 6.3 p.p., and add about 1.3 million jobs, with task creation outweighing automation.

In the context of accelerating digital transformation, quantification of the impact of artificial intelligence (AI) on the structural dynamics of resource–dependent economies is becoming particularly relevant. The purpose of the study is to develop and test an integrated dynamic model to quantify the impact of AI on Kazakhstan's structural transformation and diversification. The methodological framework includes the integration of the Bass model of innovation diffusion, an expanded production function with endogenous technological progress and the task-oriented Acemoglu–Restrepo approach, as well as a multi-criteria system of industry prioritisation. The empirical basis was based on industry data from the Bureau of National Statistics of the Republic of Kazakhstan for 2020-2024. The simulation results show that for the period 2025-2035. The cumulative increase in gross value added in the analysed industries will amount to 35.3 p.p., of which 16.8 percentage points are attributable to AI. The level of AI adoption in priority sectors reaches 86.8–93.8 p.p by 2035, which exceeds the indicators of non-priority industries by 13-32 p.p. The share of priority industries in the GDP structure increases by 6.3 p.p, while total employment increases by 22.4 p.p (+1.3 million jobs). At the same time, across all sectors, there is a steady excess of the effect of creating new tasks over the effect of automation, reflecting the specifics of a resource-dependent economy with a shortage of qualified personnel. The results confirm the expediency of concentrating government support on a limited number of industries with the greatest potential for structural transformation.

Summary

Main Finding

An integrated dynamic model estimates that AI-driven digital transformation can materially accelerate Kazakhstan’s structural transformation over 2025–2035. Simulated outcomes show a cumulative increase in gross value added (GVA) of 35.3 percentage points in the analysed industries, of which 16.8 p.p. is directly attributable to AI. Priority sectors reach very high simulated AI adoption (86.8–93.8 p.p.) by 2035, raising their GDP share and overall employment while producing net task-creation effects that outweigh automation displacement.

Key Points

Aggregate impacts (2025–2035, simulated):
- Cumulative GVA gain in analysed industries: +35.3 percentage points.
- Share of that gain attributable to AI: +16.8 percentage points.
- Priority sectors’ AI adoption by 2035: 86.8–93.8 percentage points.
- Priority vs non-priority adoption gap: priority sectors exceed non-priority by 13–32 p.p.
- Increase in priority industries’ GDP share: +6.3 p.p.
- Employment: total employment rises by 22.4 p.p. (≈ +1.3 million jobs).
Task dynamics: Across all sectors, net task creation (new tasks enabled by AI) exceeds task displacement (automation). This reflects the economy’s resource-dependence and a shortage of qualified labor, which favors AI complementing rather than replacing human tasks in the near–medium term.
Policy-relevant result: Concentrating government support (investment, training, incentives) on a limited set of high-potential priority industries is validated by the simulations as an effective strategy for structural transformation and diversification.

Data & Methods

Data:
- Industry-level data from the Bureau of National Statistics of the Republic of Kazakhstan, covering 2020–2024.
Model components (integrated):
- Bass model of innovation diffusion to simulate AI adoption trajectories across industries.
- Expanded production function with endogenous technological progress to capture productivity effects of adoption.
- Task-oriented Acemoglu–Restrepo framework to decompose impacts into automation (task displacement) and task creation/complementarity effects.
- Multi-criteria industry prioritisation to identify sectors for targeted policy support (criteria not listed here but implied to reflect structural transformation potential).
Simulation horizon: 2025–2035. Results are scenario/simulation-based (model assumptions and parameterisations determine magnitudes).

Implications for AI Economics

Structural transformation: AI can be a significant engine of productivity-led diversification in a resource-dependent economy, both by boosting GVA in targeted sectors and shifting GDP composition toward higher-value activities.
Role of prioritisation: Targeted, sector-specific policies (financing, human capital development, regulatory support) amplify AI’s structural effects; blanket or unfocused interventions are likely less efficient.
Employment effects nuance: Contrary to common automation-only narratives, in contexts with skills shortages and resource dependence, AI may predominantly create new tasks and complement labor, producing net employment gains—though composition and skill demands will change.
Policy design: Emphasis on supply-side measures (skills upgrading, task reallocation support, sectoral investment) plus measures to accelerate safe diffusion (standards, testbeds) will be crucial to realize modeled gains.
Research directions: Quantifying distributional effects (within-region and within-occupation), robustness to alternative adoption speeds and external shocks, and firm-level heterogeneity will refine policy prescriptions and help manage transitional risks.

Limitations to note: results are model-based and sensitive to diffusion parameters, technological progress assumptions, and the prioritisation criteria; the empirical base spans a recent 2020–2024 window, so shocks or global shifts after the data window could change outcomes.

Assessment

Paper Typeother Evidence Strengthlow — Results are model-driven projections rather than estimates pinned down by exogenous variation or causal identification; outcomes depend heavily on model structure, parameter choices and calibration, with no reported causal validation or robustness to alternative identification strategies. Methods Rigormedium — The study combines established theoretical building blocks (Bass diffusion, endogenous technical change, task-based framework) and uses recent national industry data, which is methodologically substantive; but rigor is limited by likely strong structural assumptions, unclear parameter estimation/validation, aggregation at the industry level, and no reported sensitivity analyses or out-of-sample checks. SampleIndustry-level data from the Bureau of National Statistics of the Republic of Kazakhstan covering 2020–2024 (gross value added, employment and related sector indicators); model calibrated to these data and used to simulate sectoral AI adoption, GVA and employment trajectories for 2025–2035, with sectors classified by a multi-criteria prioritisation system. Themesproductivity adoption IdentificationNo quasi-experimental or exogenous variation is used; causal effects are inferred from a structural simulation model that combines a Bass diffusion specification for AI adoption, an expanded production function with endogenous technological progress and a task-based (Acemoglu–Restrepo) framework, calibrated to industry-level data for Kazakhstan (2020–2024) and projected forward to 2025–2035. GeneralizabilitySingle-country focus on a resource-dependent economy (Kazakhstan) limits applicability to diversified or advanced economies, Short historical calibration window (2020–2024) may not capture longer-term trends or structural breaks, Industry-level aggregation masks firm- and task-level heterogeneity, Projections sensitive to model assumptions about diffusion rates, productivity elasticities, and the balance between automation and task creation, Policy, institutional and international economic factors specific to Kazakhstan may not generalize elsewhere, Potential measurement error in AI adoption indicators and limited validation reported

Claims (10)

Claim	Direction	Confidence	Outcome	Details
The cumulative increase in gross value added in the analysed industries will amount to 35.3 p.p. for the period 2025-2035. Firm Productivity	positive	high	gross value added (GVA)	35.3 p.p. (cumulative GVA increase, 2025–2035) 0.12
Of the cumulative 35.3 p.p. increase in gross value added for 2025–2035, 16.8 percentage points are attributable to AI. Firm Productivity	positive	high	gross value added attributable to AI	16.8 percentage points attributable to AI 0.12
The level of AI adoption in priority sectors reaches 86.8–93.8 p.p. by 2035. Adoption Rate	positive	high	AI adoption level in priority sectors	86.8–93.8 p.p. AI adoption in priority sectors by 2035 0.12
AI adoption in priority sectors by 2035 exceeds the indicators of non-priority industries by 13–32 p.p. Adoption Rate	positive	high	difference in AI adoption between priority and non-priority industries	13–32 p.p. higher adoption in priority sectors (by 2035) 0.12
The share of priority industries in the GDP structure increases by 6.3 p.p. (over the simulation horizon). Market Structure	positive	high	share of priority industries in GDP	6.3 p.p. increase in GDP share of priority industries 0.12
Total employment increases by 22.4 p.p. (equivalent to +1.3 million jobs) over the simulation period. Employment	positive	high	total employment (jobs)	22.4 p.p. increase in total employment (+1.3 million jobs) 0.12
Across all sectors, there is a steady excess of the effect of creating new tasks over the effect of automation (i.e., task-creation effects exceed automation effects), reflecting the specifics of a resource-dependent economy with a shortage of qualified personnel. Employment	positive	high	net effect: task creation vs automation	0.12
The results confirm the expediency of concentrating government support on a limited number of industries with the greatest potential for structural transformation. Governance And Regulation	positive	high	policy expediency for targeted government support	0.06
The study's methodological framework integrates the Bass model of innovation diffusion, an expanded production function with endogenous technological progress and the task-oriented Acemoglu–Restrepo approach, plus a multi-criteria system of industry prioritisation. Other	null_result	high	methodological framework (model components)	0.2
The empirical basis of the study is industry data from the Bureau of National Statistics of the Republic of Kazakhstan for 2020–2024. Other	null_result	high	data source and period used	industry data, Bureau of National Statistics of the Republic of Kazakhstan, 2020–2024 0.2