Uzbekistan's AI Strategy to 2030: What Businesses Should Know

On 14 October 2024, Presidential Resolution PP-358 approved Uzbekistan's Strategy for the Development of Artificial Intelligence Technologies until 2030. For business, this is a signal: the state treats AI as a priority and will develop infrastructure, talent and regulation while introducing AI into public services and economic sectors. Companies that start working with AI systematically now will be better placed to benefit.

What the strategy is and why it matters

A strategy is a document that sets the direction of state policy for several years. It does not oblige every company to adopt AI, but it defines where state effort will go: training specialists, computing infrastructure, sector pilots and data rules. It builds on earlier decisions — measures to create conditions for faster AI adoption were already adopted in 2021.

We will not retell the document point by point — the full text is available in the national legislation database lex.uz. What matters more is what it means in practice.

Key directions in plain words

• AI in public administration and services. Automated handling of requests, analytics, digital services for citizens. More in "AI in government organizations: is it safe?".
• Sector adoption. Healthcare, agriculture, energy, finance, education and other fields where AI improves efficiency.
• Talent. Training AI and data specialists, developing education programmes.
• Infrastructure. Data centres and computing capacity inside the country.
• Regulation and ethics. Rules for safe and responsible AI use, data protection.
• Language technology. Developing solutions that work with the Uzbek language.

The strategy is not a law that forces businesses to adopt AI. It is a map of where state resources and attention will go. Those who read the map get to the right place sooner.

What opportunities it opens for business

1. Public demand for AI solutions. Agencies will need contractors for development, integration, interfaces and staff training.
2. Local infrastructure. More data centres in the country make it easier to handle data that must still be stored in Uzbekistan by law (biometric and genetic data, telecom subscriber data) and information organizations prefer not to send abroad.
3. Talent. Over time, the market will have more specialists who can work with AI.
4. Uzbek-language products. Chatbots, voice assistants and customer services in the audience's native language are a growing niche.
5. Competitive advantage. Companies that build AI-enabled processes early will work faster and at lower cost.

Why the Uzbek language matters

Most leading models were trained mainly on English, and their quality in Uzbek lags noticeably. For a country where most people communicate in Uzbek, that is both a problem and an opportunity. Services that genuinely understand Uzbek — Latin and Cyrillic script, colloquial forms and mixing with Russian — will be in demand in public services, banking, retail and education. Companies that build up high-quality Uzbek texts and dialogues (while respecting data requirements) gain a real advantage.

What is changing in requirements

The wider AI is used, the closer the state will look at data. The Law "On Personal Data" already applies: in 2026 the requirement to store all citizens' data on servers in Uzbekistan was relaxed, but biometric and genetic data and telecom subscriber data must still be localised, and other data may go abroad only if the law's conditions are met — details in "Uzbekistan's personal data law and AI". It is reasonable to expect requirements on transparency, security and accountability in AI use to evolve.

How to read the strategy and track changes

Strategy documents usually consist of a general part setting out goals and annexes with specific measures. For business, the annexes are the most practical part to read:

• The list of measures and deadlines. It shows which projects are planned for launch and in which years.
• Responsible agencies. It makes clear who will be the customer or regulator in a given sector.
• Target indicators. They hint at which results will matter to public-sector customers.

Changes are easy to follow through lex.uz and the official websites of the relevant agencies: strategies are often supplemented by separate resolutions and roadmaps.

How business can prepare: a practical plan

1. Find 2–3 routine-heavy processes. Customer replies, document preparation, content, analytics.
2. Run a pilot. For example, an AI chatbot for customers or an AI assistant for the sales team.
3. Get your data in order. AI is only as useful as the data it works with.
4. Approve internal rules. Which services are allowed and what data may go into them.
5. Train the team. Not just "how to write prompts" but how to check answers.
6. Track regulation. New resolutions and legal changes are published on lex.uz.

How to choose your first pilot

A successful first project builds trust in AI inside the company; a failed one undermines it for a long time. Assess candidates against four criteria:

• Frequency. The task recurs daily or weekly, not once a year.
• Measurability. You can count time, the number of requests or errors before and after.
• Data sensitivity. To start with, tasks involving public or anonymised data are better.
• Cost of error. If the AI gets it wrong, the mistake can be fixed and will not hurt a customer or your reputation.

For example, drafting replies to common customer questions or writing product descriptions usually passes all four tests. Automated decisions on loans or hiring do not: both data sensitivity and the cost of error are high there.

What this means for different types of companies

• IT companies and startups — opportunities to build Uzbek-language products and take part in sector pilots and public projects.
• Retail and services — customer chatbots, personalised offers, demand analytics.
• Manufacturing and agriculture — forecasting, quality control, logistics and resource optimisation.
• Banks and financial companies — scoring, fraud prevention, customer service — with stricter data protection requirements.
• Creative and marketing agencies — faster production of content and presentations while keeping quality control and copyright in check.

If you want to work with the state as a contractor

Growing interest in AI among agencies means demand for contractors, but public-sector customers have their own expectations. It is best to prepare for them in advance:

1. Flexible data hosting. The solution must be able to run on servers in Uzbekistan or inside the customer's perimeter — this matters for biometric and other data with mandatory localisation and for government customers' official information.
2. Documentation. A description of the architecture, security measures, access procedures and data storage will be required during approvals.
3. Languages. An interface and answers in Uzbek (Latin and Cyrillic) and Russian are a requirement, not an option.
4. Transparency. The customer needs to understand how the system arrives at an answer and who checks the result.
5. Procurement procedures. Participation goes through official procedures, so learn their order and the requirements for participants in advance.

How to assess your company's AI readiness

Answer five questions honestly:

1. Do we have a task where it is clear how to measure the effect of AI?
2. Do we know where our data is stored and what state it is in?
3. Is there a person responsible for adoption and rules?
4. Is the team ready to change familiar processes?
5. Do we understand which data must not go to external services?

If the answer to most of these is "no", start not by buying technology but by putting your data and processes in order.

What to avoid

• Adopting AI "because it is trendy" without a clear task and metric.
• Uploading customer databases to foreign services.
• Expecting AI to replace staff right away — first it augments them.
• Buying expensive solutions before piloting on a small task.

AI is also changing how companies communicate with customers: websites, interfaces, presentations. If you need a clear interface for an AI service or communication that explains your product, take a look at our projects or get in touch with the Dasta Creative team.