AI as Geopolitical Power

by | Nov 25, 2025 | Uncategorized | 0 comments

Executive Summary

Russia’s characterization of AI as comparable to nuclear weapons in geopolitical significance highlights the emerging digital sovereignty race. This case study examines the implications for Singapore, a technologically advanced nation dependent on international partnerships and digital infrastructure.

Case Study: Russia’s AI Sovereignty Strategy

Background Context

Key Player: Alexander Vedyakhin, First Deputy CEO of Sberbank Date: November 2025 Forum: AI Journey event, Moscow

Core Argument

Russia positions AI development as creating a new “nuclear club” where nations either possess indigenous large language models (LLMs) or face strategic disadvantage. This represents a fundamental shift in how technological capability translates to geopolitical power.

Strategic Rationale

Data Sovereignty Concerns

  • Prohibition on uploading confidential information to foreign models
  • Requirements for domestic models in public services, healthcare, and education
  • Presidential endorsement linking AI to national sovereignty

Resource Challenges

  • Western sanctions limiting access to advanced computing hardware
  • Acknowledged gap with U.S. and Chinese competitors
  • Growing disparity in computational power

Investment Philosophy

  • Caution against “AI bubble” and excessive infrastructure spending
  • Focus on returns on investment given rapid technological change
  • Claim of immunity from overinvestment due to resource constraints

Key Players in Russian AI Development

  • Sberbank: Traditional bank transformed into AI-focused technology conglomerate
  • Yandex: Leading Russian technology firm
  • Both organizations driving national catch-up efforts

Global Outlook: The Emerging AI Order

The New Digital Divide

The AI landscape is crystallizing into distinct tiers:

Tier 1: AI Superpowers

  • United States (OpenAI, Anthropic, Google, Microsoft)
  • China (Baidu, Alibaba, Tencent, DeepSeek)
  • Characterized by massive computational resources, talent pools, and capital

Tier 2: AI-Capable Nations

  • European Union (Mistral AI, national initiatives)
  • Russia (Sberbank, Yandex)
  • Japan, South Korea, UK
  • Possess indigenous models but face resource constraints

Tier 3: AI-Dependent Nations

  • Most countries globally
  • Reliant on foreign AI infrastructure
  • Limited indigenous development capacity

Geopolitical Implications

Technological Sovereignty vs. Efficiency The tension between self-reliance and accessing best-in-class technology will define national strategies. Nations must balance:

  • Security and data sovereignty
  • Economic efficiency and innovation speed
  • International collaboration vs. independence

The Sanctions Factor Russia’s experience demonstrates how technology restrictions can accelerate fragmentation:

  • Forced development of parallel ecosystems
  • Reduced interoperability and standards
  • Increased costs and duplicated efforts

The Energy-AI Nexus Vedyakhin’s warning about energy consumption and ROI reflects growing concerns:

  • AI training and inference require massive energy
  • Environmental sustainability questions
  • Economic viability of scaling AI infrastructure

Market Dynamics

Concentration Risks

  • Dominant players may establish lock-in effects
  • Smaller nations face dependency on foreign providers
  • Potential for AI to exacerbate global inequality

Innovation Patterns

  • Open-source movements (Meta’s Llama, Mistral) challenge proprietary models
  • Specialized models for specific languages/domains gain traction
  • Edge computing and efficiency innovations reduce computational barriers

Singapore Impact Analysis

Current Position: Strengths and Vulnerabilities

Strengths

  • Advanced digital infrastructure and Smart Nation initiatives
  • Strong government capability and strategic planning
  • Highly educated workforce and research institutions
  • Strategic location and neutrality in global partnerships
  • Significant financial resources for technology investment

Vulnerabilities

  • Small domestic market limiting scale advantages
  • Dependence on imported technology and expertise
  • Limited indigenous AI model development
  • Heavy reliance on international data flows
  • Energy constraints for large-scale AI infrastructure

Strategic Dependencies

Data Sovereignty Challenges Singapore’s government services, healthcare systems, and critical infrastructure increasingly utilize cloud services and AI:

  • GovTech initiatives may rely on foreign AI providers
  • Healthcare data processed by international systems
  • Financial services dependent on global AI platforms
  • Smart city infrastructure using various technology providers

Economic Exposure

  • Finance sector dependent on AI for trading, risk management, compliance
  • Logistics and port operations utilizing international AI systems
  • Manufacturing and industrial AI largely from foreign providers
  • Professional services increasingly AI-enabled

Talent Competition Small population means competition with larger nations for AI expertise:

  • Brain drain to higher-paying U.S. and Chinese tech firms
  • Need to import foreign talent
  • Limited ability to scale indigenous research

Potential Risks for Singapore

1. Technology Decoupling If the U.S.-China AI rivalry intensifies:

  • Forced choice between technological ecosystems
  • Fragmented standards and interoperability issues
  • Reduced access to best-in-class AI capabilities
  • Increased costs from maintaining multiple systems

2. Data Localization Pressures Growing global trend toward data sovereignty:

  • Requirements to use local AI models for sensitive data
  • Higher costs for maintaining separate systems
  • Reduced efficiency from inability to leverage global models

3. Strategic Coercion Dependence on foreign AI creates vulnerability:

  • Potential service denial during geopolitical tensions
  • Surveillance concerns through foreign AI systems
  • Influence through control of critical infrastructure

4. Economic Competitiveness Nations with indigenous AI capabilities may gain advantages:

  • Faster innovation cycles
  • Better-customized solutions for local contexts
  • Economic benefits from technology exports
  • Strategic industries protected by local AI

Solutions and Recommendations for Singapore

Strategic Framework: Smart Hedging

Singapore should adopt a multi-layered approach combining selective sovereignty with pragmatic partnerships.

Solution 1: Selective AI Sovereignty

Objective: Develop indigenous capabilities for critical domains while accepting dependency in less sensitive areas.

Priority Areas for Local Development

  1. Government Services AI: Models handling citizen data, policy analysis, administrative functions
  2. Healthcare AI: Diagnostic systems, medical records analysis, public health surveillance
  3. Security and Defense AI: Intelligence analysis, cybersecurity, critical infrastructure protection
  4. Financial Regulation AI: Risk monitoring, compliance, fraud detection for regulatory purposes

Implementation Approach

  • Partner with local universities (NUS, NTU, SUTD) to develop specialized models
  • Establish national AI research consortium with sustained funding
  • Leverage Singapore’s multilingual character (English, Mandarin, Malay, Tamil) as differentiation
  • Focus on efficient, smaller models optimized for specific use cases rather than competing on scale

Estimated Investment: $500M – $1B over 5 years for critical domain models

Solution 2: Strategic AI Alliances

Objective: Build redundancy through diversified partnerships avoiding single-point dependencies.

Multi-Alignment Strategy

  • U.S. Partnership: Continue collaboration with American tech firms (Google, Microsoft, Anthropic, OpenAI)
  • European Engagement: Develop relationships with EU AI initiatives (Mistral, national champions)
  • Regional Leadership: Lead ASEAN AI cooperation framework
  • Selective China Engagement: Maintain commercial relationships while managing security risks
  • Israel and Korea: Partner with mid-sized AI innovators

Implementation Mechanisms

  • Bilateral AI cooperation agreements with multiple nations
  • Joint research initiatives and talent exchange programs
  • Regulatory sandbox for testing different AI systems
  • Procurement policies encouraging multi-vendor approaches

Solution 3: Open-Source and Open Standards Advocacy

Objective: Promote global AI ecosystem that reduces dependency on any single provider.

Key Initiatives

  • Champion open-source AI development and funding
  • Support international AI standards through ISO, ITU, and other bodies
  • Invest in open-source AI projects as public good
  • Create Singapore AI Commons for sharing government-funded models
  • Advocate for interoperability standards preventing lock-in

Benefits

  • Reduces dependency on proprietary systems
  • Enables local customization and auditing
  • Supports competitive AI ecosystem
  • Provides fallback options if commercial relationships sour

Solution 4: AI Infrastructure Resilience

Objective: Build redundant computational and data infrastructure.

Components

Distributed Computing Network

  • Establish regional data centers across ASEAN
  • Partner with neighbors for computational resource sharing
  • Invest in edge computing capabilities
  • Develop energy-efficient AI infrastructure given land constraints

Green AI Initiative Address Vedyakhin’s energy consumption concerns:

  • Focus on efficient model architectures
  • Leverage Singapore’s tropical climate for innovative cooling
  • Solar and renewable energy integration
  • Position as leader in sustainable AI development

Undersea Cable Redundancy

  • Invest in diverse submarine cable systems
  • Ensure multiple routes for data transmission
  • Reduce single points of failure in connectivity

Solution 5: Regulatory and Policy Framework

Objective: Create adaptive governance enabling innovation while managing risks.

AI Governance Framework

  • Risk-based approach: Different requirements for high-risk vs. general-purpose AI
  • Mandatory auditing for AI systems handling sensitive government/citizen data
  • Transparency requirements for AI decision-making in critical domains
  • Data localization requirements for specific sensitive categories

Cross-Border Data Flows

  • Negotiate Digital Economy Agreements (DEAs) ensuring data flow rights
  • Establish adequacy agreements for data protection with key partners
  • Create legal frameworks for emergency data access during disruptions

AI Testing and Certification

  • Establish national AI testing facility for security and performance validation
  • Certification scheme for AI systems used in government and critical sectors
  • Regular auditing of foreign AI systems for backdoors and vulnerabilities

Solution 6: Talent and Capability Development

Objective: Build sustainable local AI expertise and reduce dependence on imported talent.

Education Pipeline

  • Mandatory AI literacy in schools from primary level
  • Expanded university programs in AI and machine learning
  • Executive education for government officials and business leaders
  • Public understanding campaigns on AI implications

Talent Retention and Attraction

  • Competitive compensation for AI researchers and engineers
  • Immigration pathways for AI talent
  • Research grants and startup support
  • Quality of life advantages for international talent

Industry-Academia Collaboration

  • Joint appointments between universities and industry
  • Sabbatical programs for academics to work in industry
  • Industry-sponsored research projects
  • Graduate employment partnerships

Solution 7: Economic Positioning

Objective: Turn AI challenges into economic opportunities.

AI Services Hub Position Singapore as:

  • Neutral testing ground for AI systems from different providers
  • AI compliance and auditing center for Asia-Pacific
  • Regional AI training and deployment hub
  • Trusted intermediary for AI governance

Specialized AI Applications Develop expertise in:

  • Tropical urban AI (climate, density, sustainability)
  • Multilingual and multicultural AI systems
  • Maritime and logistics AI
  • Financial services AI
  • Smart city AI

AI Export Potential

  • Package Singapore’s AI governance model for other small nations
  • Export specialized AI applications to region
  • AI consulting and integration services
  • AI security and auditing services

Implementation Roadmap

Phase 1: Foundation (2025-2027)

Year 1 Actions

  • Conduct comprehensive AI dependency audit across government and critical sectors
  • Establish National AI Sovereignty Task Force with ministerial authority
  • Launch priority research programs in critical AI domains
  • Initiate bilateral AI partnership discussions with 5-7 key nations
  • Begin AI literacy programs in schools and government

Year 2 Actions

  • Release first Singapore-developed AI models for non-sensitive applications
  • Complete AI infrastructure resilience assessment
  • Establish AI testing and certification facility
  • Launch regional ASEAN AI cooperation initiative
  • Implement enhanced procurement rules for AI systems

Phase 2: Build-Out (2027-2030)

  • Deploy locally-developed AI for critical government services
  • Establish Singapore as regional AI governance hub
  • Create operational distributed computing network across ASEAN
  • Achieve 50% redundancy in critical AI applications
  • Launch AI export services program

Phase 3: Leadership (2030+)

  • Position as global model for small nation AI strategy
  • Lead international AI standards development
  • Establish self-sufficiency in critical AI domains
  • Become net exporter of specialized AI capabilities

Success Metrics

Security Metrics

  • Percentage of critical systems with indigenous or diversified AI: Target 80% by 2030
  • Number of strategic AI partnerships: Target 10+ by 2027
  • Data sovereignty compliance rate: Target 100% for sensitive categories

Economic Metrics

  • AI sector contribution to GDP: Target 5% by 2030
  • AI-related jobs created: Target 50,000 by 2030
  • AI services exports: Target $5B annually by 2030

Capability Metrics

  • Local AI talent pool size: Target 20,000 AI professionals by 2030
  • Number of locally-developed AI models: Target 15+ specialized models by 2030
  • AI research publications: Top 20 globally by 2030

Conclusion

Russia’s framing of AI as the new nuclear weapon highlights the strategic significance of technological sovereignty. For Singapore, this presents both challenges and opportunities.

The nation cannot and should not attempt to compete directly with AI superpowers in computational scale or general-purpose model development. Instead, Singapore should pursue strategic selectivity: developing indigenous capabilities for critical domains while maintaining diversified international partnerships for general applications.

Success requires balancing three imperatives:

  1. Security: Ensuring critical systems are not dependent on single foreign providers
  2. Efficiency: Accessing best-in-class AI capabilities for economic competitiveness
  3. Leadership: Positioning as a trusted, neutral hub for AI governance and innovation

By leveraging its strengths in governance, education, infrastructure, and strategic positioning, Singapore can navigate the AI geopolitics landscape while turning potential vulnerabilities into competitive advantages. The key is acting decisively now, before technological dependencies become entrenched and options narrow.

The emerging AI order will reshape global power dynamics. Singapore’s response will determine whether it remains a relevant player in the digital age or becomes a passive consumer of foreign technology with all the vulnerabilities that entails.