Kazakhstan’s Kurultai Reform: How AI and Digital Governance Are Reshaping Central Asia

Kazakhstan has very recently completed a major constitutional reform. On 23 August 2026, the Central Asian state held the first election of deputies to the Kurultai, the country’s new unicameral Parliament, replacing the bicameral structure of the Senate and Mazhilis that had governed the legislative process since independence. The Kurultai’s establishment traces back to the constitutional referendum of 15 March 2026, itself the latest step in a reform trajectory that Astana has pursued since the January 2022 unrest, when President Kassym-Jomart Tokayev launched what has come to be known domestically as the transition toward a “New Kazakhstan.” That trajectory has combined constitutional amendments limiting presidential power and now a redesigned legislature, with the government ultimately aiming for a shift from a super-presidential model toward a more balanced distribution of political authority. The results of this election provide an early indication of how the reformed institutional structure will translate into practice and promises to be impactful across several directions.

Prospects for the Reform: Institutional Design and Legislative Capacity

The shift from a bicameral to a unicameral legislature is, in institutional design terms, a wager on speed. The legislative process is expected to have fewer veto points, and this should allow Kazakhstan to move more efficiently on policy areas that require sustained, technically demanding legislative attention, such as digital transformation and AI governance, where Kazakhstan, as the only Central Asian country with AI legislation, can share best practices and lead the region. A single 145-seat body elected through nationwide party-list proportional representation comes to replace the bicameral system, and this can result in more efficient or faster governance if handled correctly.

The election also had a new institutional commitment, as lists were required to reserve a combined share of at least 30% for women, young people, and persons with disabilities. This was a tangible move towards inclusivity, being supported also by the participating parties that had 172 women, 74 candidates under the age of 35, and 19 candidates with disabilities against 545 total registrations. Few electoral systems in the region write inclusion quotas of this kind directly into candidate list requirements, and the fact that parties exceeded the threshold indicates that the provision was not just a formality.

Digital Transformation and the Governance of AI

Kazakhstan’s institutional redesign arrives at a moment when the country has moved from strategic intent to binding regulation faster than any other state in the region. This comes at a time when Kazakhstan is trying to lead other major changes, such as on digital transformation. A prime example is the law on artificial intelligence, which has been in force ever since January 2026. This law made Kazakhstan the second jurisdiction globally, after the European Union, to adopt a standalone AI law, built on a risk-based classification system that draws structurally on the EU’s own tiered approach while adapting it to domestic priorities. The additional creation of a dedicated Ministry of Artificial Intelligence and Digital Development in September 2025, paired with the new unicameral Kurultai, gives a unique opportunity to Kazakhstan, since a single legislative chamber working alongside a single AI ministry reduces the coordination friction that fragments AI governance efforts in many larger states. Fewer structural veto points should, in principle, let Astana sustain the kind of continuous legislative attention that digital and AI governance demands, since these are policy domains that rarely resolve through a single bill and instead require iterative regulatory adjustment as technology and international standards evolve.

Reflecting on what Kazakhstan has built so far, it sets the pace for a rather ambitious future. The law’s risk-based structure, its mandatory labeling requirements for synthetic content, and the planned National AI Platform, anchored by the $10 billion Data Center Valley project with Firebird and NVIDIA, together signal that Astana is treating AI as a state-building priority. At the same time, key implementation pieces remain in progress. The data governance requirements that will determine how the national platform’s data libraries are populated and secured have not yet been finalized, and the unified Digital Qazaqstan strategy meant to consolidate the country’s various digital initiatives into a single coherent framework is still being drafted. This pace of events reflects how every jurisdiction has transitioned from AI legislation to AI implementation and the challenges that it has faced, especially in many EU member states. Hence, the process to close the gap between legislative architecture and operational capacity will be paramount for Kazakhstan’s AI governance trajectory. Whether the Kurultai’s structural advantages translate into governance quality will depend on the human capital and inter-agency coordination choices. Kazakhstan makes over the next phase of implementation an area where its Erasmus+ partnerships and academic cooperation with EU institutions could play a meaningful supporting role.

Kazakhstan in the C5+1 Framework: Institutional Reform and Regional Positioning in the Era of AI

Kazakhstan’s constitutional and legislative overhaul does not unfold in isolation from its external engagements, and the C5+1 format, bringing together the five Central Asian republics with the European Union, offers a useful lens for reading the reform’s regional significance. Brussels’ engagement with Central Asia through this format has initially focused on connectivity and raw materials, with larger states like France and Germany historically anchoring the agenda on these sectors. However, it has recently extended into digital and AI cooperation, and in these sectors some of the bloc’s most influential innovations in digital governance, particularly software-based, service-delivery models, emerged from Eastern European and smaller member states, highlighting again the importance of a C5+1 model that is initiated by such EU states. Estonia is a great example, having built its digital governance architecture, from X-Road data infrastructure to e-residency and digital identity systems, into a reference point cited well beyond Europe, despite being one of the EU’s smallest members by population. This shows that such a model of cooperation can be adapted by other similar EU states, Greece being among them. Several examples relevant to AI governance can be found here. A regulatory sandbox partnership can function perfectly, as Greece, like several EU states, has experience piloting and leading controlled testing environments for regulated technologies, and a joint Greek-Kazakh sandbox, in the context of C5+Greece, for high-risk AI systems under Kazakhstan’s new risk-based classification could let both sides stress-test compliance mechanisms before full-scale enforcement, benefiting from direct comparison against the EU AI Act’s own risk tiers. A second track concerns AI governance in financial and tax compliance, an area where both the EU and Kazakhstan are actively developing rules on algorithmic transparency and automated decision-making in revenue administration, and where structured exchange on audit standards, explainability requirements, and data provenance could produce transferable frameworks. Greece’s own experience is particularly relevant here, since its tax authority has developed one of the EU’s most extensive digital enforcement infrastructures, including real-time e-invoicing and algorithmic audit-risk scoring, built specifically in response to the country’s longstanding tax evasion challenge, giving it practical, tested expertise in exactly the kind of AI-driven compliance systems Kazakhstan’s own tax authorities are likely to develop. Through a Greek-Kazakh collaboration, such practices can be adapted to the Kazakh socio-cultural and economic reality. A third track lies in ESG-related AI governance, specifically the growing use of large language models to extract and standardize corporate ESG disclosures, an area of active academic work in smaller Southern and Eastern European countries like Greece and Estonia, where joint research on extraction accuracy, disclosure verification standards, and regulatory alignment with EU sustainability reporting rules could position Kazakhstan’s emerging AI governance framework alongside established EU compliance regimes rather than developing in parallel to them.

Smaller EU member states are often better positioned than larger ones to move quickly on this kind of pilot cooperation and knowledge-sharing arrangement because they carry less institutional weight to coordinate internally before acting. Kazakhstan’s own parallel bet on speed through the Kurultai reform suggests an EU partner profile that fits well with this kind of agile, bilateral track rather than only the EU’s slower-moving collective instruments. Whether channeled through Greek, Estonian, or other member state partnerships, the C5+1 format’s real value going forward may lie less in large-scale collective initiatives and more in these targeted bilateral links between Kazakhstan’s fast-moving regulatory architecture and the smaller EU states that have already demonstrated what focused, well-designed digital governance can achieve.