Georgia risks missing AI-fuelled economic gains because it has no national strategy or legal framework; adopting an integrated, adaptive policy and infrastructure program modeled on successful small-country approaches would better position it as a regional technology leader.

Main Finding

Georgia lacks a national AI strategy and legal definition for AI, producing fragmented, sectoral adoption that limits its ability to capture the sizable macroeconomic gains AI can deliver. The paper argues that, based on international practice, Georgia should adopt an integrated strategic framework—combining regulation, infrastructure investment, education/R&D support, ethical standards, and cross‑sector coordination—to position itself as a regional AI leader and avoid falling further behind in the global, uneven distribution of AI benefits.

Key Points

• Macroeconomic forecasts:
  • Goldman Sachs: generative AI could raise global GDP ~7% over a decade (~$7 trillion) via ≈1.5 pp higher labor productivity.
  • Penn Wharton: generative AI effects on labor productivity grow over time (projected ≈1.5% by 2035 → 3% by 2055 → 3.7% by 2075).
  • IMF: AI could raise global GDP ~4% over the next decade but with sharply unequal gains (developed countries likely get ~2× the benefit of low‑income countries).
  • McKinsey (2025): 88% of organizations use AI in ≥1 function; 62% working on agentic models; many remain in pilots due to data and talent gaps.
• Technological & regulatory context (circa 2026):
  • Agentic AI and advanced LLMs (e.g., GPT‑5.4, Claude 4.6) mark a shift to autonomous multi‑agent systems and "Native Computer Use."
  • Three regulatory poles: EU (risk‑based ethical regulation + InvestAI infrastructure spending), US (deregulatory / sectoral approach), China (state‑centered, data/localization/sovereignty model).
  • Multilateral instruments (Council of Europe, OECD, UNESCO, 2026 UN resolution) are converging minimum AI safety norms.
• Sectoral opportunity for Georgia:
  • Finance/FinTech: high regional tech maturity in banking; AI for credit scoring, fraud detection, personalized advising tied to National Bank FinTech strategy.
  • Healthcare: AI diagnostics (medical imaging), predictive analytics to address specialist shortages in regions.
  • Education & public services: personalized learning, automation of services to reduce bureaucracy and improve human capital.
• AI nationalism and infrastructure race:
  • Massive state investments globally (US ~$500B package including “Stargate”; China large sovereignty funds; CHIPS/semiconductor drives; EU, UK, Gulf states also scaling compute and talent).
  • Semiconductor and compute policies (e.g., CHIPS Act 2.0, China Big Fund) shape who can train/host frontier models.
• Comparative policy insight:
  • Small/medium states (Singapore, UK, Canada, France highlighted) achieved disproportionate outcomes via institutional design, targeted incentives, public–private coordination, compute programs, and regulatory clarity.
  • Examples: Singapore introduced strong tax incentives and national AI strategy; UK combined sectoral institutions, AI Safety Institute, and an AI Bill; Canada combined long‑term national strategy, compute investments (Canada AI Compute), and strict AIDA regulation.

Data & Methods

• Methods: qualitative policy analysis and comparative case study informed by secondary sources and published forecasts.
• Sources synthesized: institutional forecasts/reports (Goldman Sachs, McKinsey, Penn Wharton, IMF), national strategy documents (Singapore, UK, Canada, France), national budgets and program reports (e.g., CHIPS Act, Canada Budget 2024), international indices (Government AI Readiness, Global AI Index), and multilateral instruments (OECD, UNESCO, Council of Europe, UN resolution).
• Case selection: purposeful sampling of countries with diverse geographic, political and technological profiles that have succeeded without relying exclusively on US/China models (focus on alternative “third‑way” approaches).
• Sectoral assessment for Georgia: mapping of national strategies/policies (where present) and alignment with sectoral strategies (National Bank FinTech, Digital Health) and institutional capacities.
• Limitations: relies on secondary projections and policy documents rather than original empirical fieldwork in Georgia; forecasts carry uncertainty, especially long‑run productivity estimates and scenario dependence on compute/talent access.

Implications for AI Economics

• For Georgia specifically:
  • Urgent need for a national AI strategy and legal definition to reduce fragmentation, coordinate investments, and signal incentives to private actors.
  • Priority areas: build compute capacity (or secure regional/cloud partnerships), unlock public data for safe commercial use, invest in digital skills/education, embed ethics and algorithmic‑bias controls, and align sectoral strategies (finance, health, education).
  • Low‑cost/high‑impact levers: targeted tax incentives, public–private R&D consortia, data‑sharing frameworks, pilot deployments in finance and health to demonstrate early wins.
• Broader economic implications:
  • AI promises substantial aggregate productivity and GDP gains but is likely to widen global inequality without active policy to support latecomers (risk of “digital divergence”).
  • Small and middle‑income countries can still capture disproportionate benefits by adopting targeted policies (institutional flexibility, open government data, compute partnerships, talent development) rather than attempting resource‑intensive sovereignty strategies.
  • Regulation matters: balancing innovation and safety (e.g., "high‑impact" AI rules) can coexist with competitive AI ecosystems if designed to reduce uncertainty and enable testing/standards.
  • Investments in semiconductors, cloud/compute, and human capital are strategic bottlenecks—policy should prioritize easing access to compute and strengthening AI talent pipelines to realize productivity gains.
• Governance challenge: the Agentic AI era intensifies issues around intellectual property, data privacy, content safety, and model risk—requiring coordinated national frameworks that are interoperable with international norms.