Uzbekistan Aims to Become a Regional Leader in Artificial Intelligence, but SMEs Face Significant Barriers

Tashkent, Uzbekistan (UzDaily.com) — Uzbekistan has set an ambitious goal of becoming a regional hub for artificial intelligence by 2030, yet small and medium-sized enterprises across the country face serious obstacles in adopting AI technologies, according to a new joint report by the UNDP and the Ministry of Economy and Finance entitled “AI Adoption in Uzbekistan’s Private Sector: Drivers, Challenges and Recommendations.”

Under the National Strategy for the Development of Artificial Intelligence Technologies through 2030, approved by a presidential decree in October 2024, Uzbekistan aims to generate US$1.5 billion annually in AI-based products and services, integrate AI into 10 percent of e-government services, and rank among the world’s top 50 countries in the AI Readiness Index published by Oxford Insights.

Small and medium-sized enterprises account for around 92 percent of all companies in Uzbekistan, employ roughly three quarters of the workforce and generate more than half of GDP. In the first three quarters of 2024 alone, small businesses contributed 53.8 percent of national GDP. Despite their central role in the economy, these firms face structural constraints that limit their ability to fully benefit from artificial intelligence.

The study, conducted under a joint UNDP–Ministry of Economy and Finance project aimed at empowering youth for the development of the digital economy and digital entrepreneurship, identifies four main categories of barriers to AI adoption by SMEs.

Financial constraints remain among the most acute. Venture capital investment in Uzbekistan increased from US$0.3 million in 2020 to US$69.5 million in 2024, representing more than 230-fold growth. Even so, these figures remain modest compared to regional peers. In the same year, Kazakhstan attracted US$71 million in venture funding, Türkiye US$1.1 billion, and India US$13.7 billion.

Outdated corporate and investment legislation further exacerbates the problem. Current regulations do not recognize standard venture instruments such as SAFE agreements, convertible notes or preferred shares, preventing investors from formally registering equity stakes. Commercial banks continue to rely on collateral-based lending models that are ill-suited to AI startups, which often lack physical assets but possess valuable intellectual property.

Infrastructure and technical challenges form another major barrier. High computing costs and dependence on foreign cloud providers place heavy financial pressure on startups. One founder developing natural language processing tools for Uzbek and regional languages reported spending between US$22,000 and US$25,000 per month on cloud computing, with additional peak costs of up to US$6,000 for model training. Delays in deploying a national cloud platform and high-performance computing infrastructure further restrict access to affordable computing resources. Uzbekistan also suffers from a severe shortage of high-quality Uzbek-language text corpora, labeled datasets and standardized benchmarks, posing particular challenges for NLP, speech recognition and speech synthesis systems.

The shortage of AI talent represents the third major obstacle. Only a limited number of universities currently offer specialized AI programs or tracks in machine learning, natural language processing or computer vision. As a result, most AI startups must train junior specialists internally for three to six months before they can contribute meaningfully to development. While the national AI strategy envisions the creation of 10 AI laboratories and the training of 1,000 highly qualified AI specialists, with pilot labs already launched at INHA University and TUIT, these initiatives remain at an early stage and require sustained investment and closer industry collaboration.

Data governance and AI regulation constitute the fourth group of barriers. Uzbekistan’s 2019 Personal Data Law established a basic legal framework for privacy and security, but compliance requirements such as mandatory database registration and data localization impose disproportionate costs on small firms. Although the national AI strategy outlines principles of fairness, transparency and human oversight, these remain aspirational and have not yet been translated into binding, auditable requirements such as standardized documentation, explainability protocols or evaluation benchmarks.

In April 2025, the Legislative Chamber of the Oliy Majlis adopted at first reading a draft law “On the Regulation of Relations Arising from the Use of Artificial Intelligence.” The bill seeks to establish safeguards to protect citizens from harm caused by irresponsible AI use and предусматривает mandatory labeling of AI-generated content.

The report recommends a phased, proportionate approach that clearly distinguishes the roles of regulators, market intermediaries and market participants. For regulators, it suggests adopting a risk-based, staged framework aligned with international practice, starting with sector-specific guidance and soft-law instruments before moving toward comprehensive AI legislation. It also proposes introducing AI and data maturity assessment frameworks for key agencies, developing AI-specific procurement guidance and simplifying privacy rules while strengthening structured data-sharing mechanisms.

For market intermediaries such as IT Park Uzbekistan, academic institutions and venture funds, the recommendations include expanding applied AI education and scholarship programs, particularly outside Tashkent, diversifying financing and investment instruments, and strengthening support for AI community networks. Market participants are encouraged to accelerate AI experimentation among startups, SMEs and industry, establish safe data-sharing mechanisms and actively engage in policy consultations and standard-setting processes.

Uzbekistan has made notable progress in AI readiness. In the GAIR 2024 index, the country scored 53.45 points, ranking 70th globally and third in South and Central Asia, ahead of all other Central Asian states. This reflects strong strategic direction while also highlighting gaps in infrastructure, data governance and implementation capacity.

The country’s AI ecosystem is already delivering tangible results in priority sectors. In agriculture, startups Fermerly and Growz provide AI-based agronomy recommendations and plant disease diagnostics. In healthcare, Sino AI has developed a digital AI-powered medical assistant, while AIScan applies machine learning algorithms to CT and MRI image analysis. In the financial sector, Uzum, Uzbekistan’s first technology unicorn valued at approximately US$1.16 billion, is integrating AI across its fintech and e-commerce platforms.

The government is also embedding AI into public service delivery. Uzinfocom, the national systems integrator, operates MyID, a digital identification platform processing millions of transactions daily, and has developed Muxlisa AI, a national natural language processing system trained on a five-billion-word Uzbek-language corpus. The Ministry of Justice has launched an AI-powered chatbot on the Lex.uz platform to handle legal inquiries.

An annex to the report examines the use of regulatory sandboxes for AI. By February 2025, the Datasphere Initiative had identified 66 data and AI sandboxes across 44 countries, underscoring growing interest in sandbox-based experimentation to balance innovation with safeguards. While Uzbekistan has multiple sandbox-style initiatives in blockchain, cryptoassets, capital markets and digital technologies, none are specifically designed for AI. Stakeholders highlighted the need for a dedicated AI sandbox to address issues such as data access, model validation and algorithmic accountability.

The proposed AI sandbox framework for Uzbekistan identifies high-potential areas aligned with national AI strategy goals and SME needs, particularly in healthcare, finance and banking, and energy. It recommends clearly defining the sandbox’s problem scope and objectives, assessing the suitability and institutional readiness of lead agencies, establishing governance and stakeholder structures, and designing transparent eligibility criteria that prioritize SMEs.

Akiko Fujii, UNDP Resident Representative in Uzbekistan, noted that artificial intelligence is reshaping economies and societies worldwide, affecting business models, education, public services, agriculture and healthcare. The UNCTAD Technology and Innovation Report 2025 highlights AI’s transformative potential while warning of widening inequality if countries fail to accelerate inclusive technological development.

Ilkhom Norqulov, First Deputy Minister of Economy and Finance of Uzbekistan, emphasized that while ongoing improvements in digital infrastructure and regulatory reforms provide an important foundation, the report’s findings show that long-term success will depend on creating a predictable, dynamic and innovation-friendly environment that enables AI entrepreneurs to confidently scale their ideas and contribute to national economic development.