• KZT/USD = 0.00214
  • TJS/USD = 0.10810
  • UZS/USD = 0.00008
  • TMT/USD = 0.29850
  • KZT/USD = 0.00214
  • TJS/USD = 0.10810
  • UZS/USD = 0.00008
  • TMT/USD = 0.29850
  • KZT/USD = 0.00214
  • TJS/USD = 0.10810
  • UZS/USD = 0.00008
  • TMT/USD = 0.29850
  • KZT/USD = 0.00214
  • TJS/USD = 0.10810
  • UZS/USD = 0.00008
  • TMT/USD = 0.29850
  • KZT/USD = 0.00214
  • TJS/USD = 0.10810
  • UZS/USD = 0.00008
  • TMT/USD = 0.29850
  • KZT/USD = 0.00214
  • TJS/USD = 0.10810
  • UZS/USD = 0.00008
  • TMT/USD = 0.29850
  • KZT/USD = 0.00214
  • TJS/USD = 0.10810
  • UZS/USD = 0.00008
  • TMT/USD = 0.29850
  • KZT/USD = 0.00214
  • TJS/USD = 0.10810
  • UZS/USD = 0.00008
  • TMT/USD = 0.29850
11 August 2026
11 August 2026

AI E-Government in Kazakhstan Expands Into Jobs and Benefits

@depositphotos

Kazakhstan is preparing a deeper overhaul of its e-government system, with artificial intelligence expected to identify which service citizens need, match unemployed people with vacancies, and assist officials in identifying potential recipients of social assistance.

The plans were outlined at a government meeting on August 11 by Minister of AI and Digital Development Zhaslan Madiyev and Minister of Labor and Social Protection Askarbek Yertayev. Several are expected to expand nationwide by the end of 2026.

eGov GPT, which the government was testing earlier this year, currently works with 50 of the most frequently used public services. According to Madiyev, these account for around 30% of citizens’ requests. The government plans to increase that share to 70% by the end of the year. Madiyev said the assistant was used more than 150,000 times during its first four months. An official government description of the system says it can identify the appropriate public service from a user’s request and help complete it online.

The idea is for users to simply describe their situation in ordinary language, for example, “I need to register my address” or “My child is starting school this year.” eGov GPT identifies the relevant procedure, helps prepare the application, and guides it through the government system.

Kazakhstan began building its e-government system in the 2000s, long before the current boom in generative AI. In the latest UN E-Government Survey, the country ranked 24th among 193 states, the highest position in Central Asia. The index assesses countries’ online public services alongside the infrastructure and human capacity needed to use them.

What is changing now is the way citizens interact with the state. Previously, digitizing a service meant that instead of travelling to a Public Service Centre, Kazakhstan’s equivalent of a government service centre, a citizen could fill in the required form online. The new model is intended to remove part of that work as well.

According to Madiyev, the number of requests handled directly by Public Service Centre staff has already fallen by around 30%. The authorities plan gradually to transform the centres into what they call AI offices. Visitors will be able to submit a request through eGov GPT, apply for an identity card, passport, or driving licence through a terminal, and collect the finished document from a parcel locker, including at weekends.

The government has not said what the shift will mean for staffing levels at the centres. Nevertheless, authorities plan to go considerably further in the labor market.

A new Digital Employment Service will compare government-held information on an unemployed person’s occupation, qualifications, and skills with employers’ requirements. The Labor Ministry describes the idea as moving towards a model in which “the person does not look for a job, but the job finds the person”.

When a match is found, the citizen will receive an offer through eGov, the state employment portal Enbek.kz, or an SMS from the 1414 service number. The system will also be able to recommend training or retraining, or suggest relocation if a suitable vacancy is available in another region.

The first trial took place in July in the Pavlodar region in north-eastern Kazakhstan. Potential jobseekers were sent 39,000 messages. According to Yertayev, 962 recipients subsequently found work. However, the figures presented by the minister do not show how many of those people obtained jobs as a direct result of the new service.

The service is planned to launch in the western Mangystau region in September and expand nationwide by the end of the year.

From October, algorithms are expected to be introduced into a more sensitive area: social assistance. AI is intended to identify potential recipients of state payments in advance, using information from government databases.

The idea of identifying potential beneficiaries automatically is not entirely new. Kazakhstan’s Digital Family Card has operated since 2022, using information held by state agencies to identify citizens who may qualify for support and contact them proactively. The new system is intended to expand the role played by AI in that process.

Citizens will need to submit only one application. The system will collect the required information and prepare a draft decision on the payment. The final decision, however, will remain with the relevant local government body.

That distinction is important. An error by a chatbot suggesting the wrong public service and an algorithmic error in determining whether someone qualifies for a benefit have very different consequences. In the case of social payments, AI is therefore being assigned the role of an assistant rather than the final decision-maker.

Kazakhstan is experimenting with algorithms in other government services as well. The Times of Central Asia previously reported on plans to use AI to monitor driving licence examinations, as well as the creation of a fund to finance artificial intelligence projects with the participation of the National Bank.

Dmitry Pokidaev

Dmitry Pokidaev

Dmitry Pokidaev is a journalist based in Astana, Kazakhstan, with experience at some of the country's top media outlets. Before his career in journalism, Pokidaev worked as an academic, teaching Russian language and literature.

View more articles fromDmitry Pokidaev

Suggested Articles

Sidebar