This article draws on two sources: official Geostat data (Use of information and communication technologies in enterprises, 2025) and BTU’s survey of 37 small and medium service-sector enterprises, July 2026. The national data shows the overall adoption picture; the BTU survey shows how use develops inside individual firms.
Scale of Adoption and Technologies Used
In 2025, only about 4% of Georgian enterprises used AI in any form. The rate varies sharply by size – large enterprises are roughly eight times more likely to use AI than small ones, pointing to a shortage of resources and technical competence among smaller firms.
Text analysis clearly dominates among AI-using enterprises, used by about two thirds of them. Generating images, video or audio comes second; autonomous machines such as robots and drones are the least used of all.
What AI Is Used For
Marketing and sales is the clear leader – a third of AI-using enterprises apply it there. Organising business administration comes second; managing the enterprise itself is the least common purpose.
40% of AI-using enterprises could not name a specific business purpose for their AI use, suggesting the technology often enters a firm before it is attached to any working process.
The tightest pairing excludes marketing entirely – it is business administration and research and development, both concerned with internal work. Marketing looks more like an entry point than a final goal.
Who Uses AI More Often
Among firms with online sales, the AI adoption rate is 16%, against 3% without – nearly a fivefold gap. Yet only about half of these firms use AI directly for marketing and sales; online trade is more a marker of digital maturity than a cause of adoption.
Big data and AI remain largely independent practices in Georgia. 43% of big-data users also use AI, but only 10% of AI users use big data – data infrastructure alone does not lead to AI adoption.
Having an ICT specialist remains the strongest predictor of AI adoption, yet 72% of AI-using firms have none at all – almost a third of that group modified or built their own AI software.
Most firms with a website chat service use no AI at all – the chat is operator-based or scripted. These firms tend to be smaller, less digitally equipped, and rarely have a technical team.
How AI Develops Inside SME Firms
AI use inside a firm follows a clear sequence. Nearly every AI-using firm starts with document preparation and data analysis. Marketing content and customer service come next, the point where AI output reaches outside the company. Finance, security monitoring and HR arrive last, only in firms already using AI broadly.
Firms at the basic level mostly describe time savings – the same work done faster. Advanced users mention this less, not because it disappears, but because larger changes take over.
Basic users almost unanimously report no revenue effect. Applied-stage firms mostly say they simply do not measure it. Only advanced firms report a modest revenue increase.
AI literacy spreads mainly through colleagues sharing experience horizontally, not training. Formal training is rare and appears almost only in advanced firms, focused on coding and safe use.
Basic and applied firms describe their AI costs as just a subscription or licence. Advanced firms report a different structure – GPU servers, model fine-tuning, API charges, and integration into their own systems.
The leading reason for not using AI is that firms see no benefit in their own case; the second is a lack of employee knowledge. These firms remain open to adopting AI in the future.
The full report – every chart, additional data, and methodology details – is available Here: “AI INTEGRATION IN BUSINESS SECTOR
IN GEORGIA ” https://btu.edu.ge/en/kvleva/btu-s-mkvlevarebi/ongoing-and-completed-research-projects/ai-ekosistemis-daijesti/
Source: Geostat, Use of information and communication technologies in enterprises, 2025; BTU survey on AI integration in business activity, July 2026. Author: Tsotne Zhghenti, BTU Affiliate Associate Professor.