2026制定人工智能政策以驱动创新洞察报告_141页_2mb

GTIPA Perspectives: Making AI Policy to Drive Innovation Summary

Core Content

The Global Trade and Innovation Policy Alliance (GTIPA) is a coalition of over 60 think tanks from 40+ countries, advocating for market-based, rules-governed trade and innovation policies to maximize global welfare. The report presents a comprehensive analysis of AI innovation policies across 30 countries, highlighting the diverse strategies and outcomes, and emphasizes the need for proactive, holistic, and context-sensitive approaches to AI governance.

Main Views

1. Proactive and Holistic Strategies

GTIPA members stress that governments must be proactive and holistic in their AI strategies. Most countries have developed coordinated national AI strategies, which are more effective when they include:

• A variety of levers to introduce and catalyze AI deployment.
• Specific plans and benchmarks for implementation.

2. Balance of Support and Freedom

Successful AI policies require a thoughtful balance between support and freedom. Governments should:

• Allow room for experimentation and innovation.
• Acknowledge the role of government as a patient capital provider, procurer, and adopter.
• Align with the Innovation Success Triangle concept, which emphasizes the importance of market freedom, investment, and innovation.

3. Context-Sensitive Policies

Each country must tailor its AI strategy to its unique context, strengths, and weaknesses. Examples include:

• Italy focusing on applied AI solutions for the real economy.
• Chile and Sweden leveraging their existing strengths in renewable energy and telecommunications.
• El Salvador prioritizing AI education and public health applications.

4. Ecosystem of Allies

AI deployment is most successful when supported by a rules-based, competitive trade and globalization framework. Countries like El Salvador, the EU, and the U.S. demonstrate the importance of collaboration and international cooperation.

Key Information

Compute and Infrastructure Capacity

Several countries are investing heavily in compute and energy infrastructure:

• Argentina partnered with OpenAI and Sur Energy to build the Stargate data center in Patagonia, expected to be the largest AI infrastructure facility in Latin America.
• Chile is developing the Humboldt submarine cable, which will connect South America to the Asia-Pacific.
• Mexico is positioning Querétaro as its leading data center hub.
• United States has eased regulations to support AI data center development.

Research and Development (R&D)

Innovation requires R&D institutions and collaborative frameworks:

• Bulgaria established the INSAIT Institute with Swiss partners.
• Greece is building the DAEDALUS supercomputer and the PHAROS AI Factory.
• Italy has the Leonardo supercomputer, providing over 500 million computing hours.
• South Africa launched the AI Institute of South Africa.
• Serbia partnered with NVIDIA to create the National AI Platform.
• United States introduced the Genesis Mission to build a network of AI labs and datasets.

Tax and Investment Ecosystems

Competitive tax policies and investment incentives are crucial for innovation:

• Argentina introduced the Incentive Regime for Large Investments (RIGI), offering tax and customs stability for 30 years.
• Australia established a $15 billion National Reconstruction Fund.
• Brazil allocated $1 billion in credit and equity for AI projects.
• Bangladesh offers a 10-year tax holiday and duty-free imports for capital machinery.

Talent and Skills

Investing in education and skills development is essential to manage AI-related fears of unemployment:

• Australia partnered with Microsoft to launch the National AI Skills Accelerator.
• Ecuador aims to train 10,000 workers.
• Pakistan launched a National AI Skill Development Program with a “Train the Trainer” initiative.
• Philippines created the Education Center for Artificial Intelligence Research (ECAIR).

Adaptive Regulation

Adaptive and flexible regulatory frameworks are vital to support innovation:

• Indonesia uses regulatory sandboxes and risk-based obligations.
• South Africa iterated through three AI policy drafts with public and international input.
• Vietnam adopts an ex-post control model with a tiered risk assessment.
• United States uses a voluntary frontier framework for AI models.

Government as Adopter

Governments can be key adopters and drivers of AI innovation:

• Albania launched Diella, an AI-driven digital assistant.
• Argentina implemented PROMETEA, reducing case resolution times by 77%.
• Brazil reports 182 AI solutions in operation.
• Chile created the Ciudades por la IA program to support local AI use.
• Colombia established guidelines for AI in the judiciary.
• El Salvador launched DoctorSV, a telemedicine platform with 2.3 million consultations in 5 months.
• Peru created an Artificial Intelligence Officer.
• Philippines uses AI in education through the System for Intelligent Growth and Learner Anthropometry (SIGLA).

Sectoral and SME Deployment

AI should be accessible to all sectors and SMEs:

• Bangladesh uses KrishokBondhu for AI-powered agricultural advice.
• Germany supports SMEs through the Mittelstand-Digital project.
• Italy helps manufacturing companies through the MADE Competence Center.
• Mexico sees Nissan's automation initiatives reduce maintenance failures.

Standards and Coordination

Collaboration and international alignment are essential for AI deployment:

• Costa Rica joined Latam-GPT and Pax Silica.
• Pakistan aligned its policy with OECD AI Principles and ITU AI for Good.
• South Africa helped develop the Artificial Intelligence for Africa Blueprint.

Appendix: A Global Declaration on Free and Open AI

The report includes a Declaration of Open AI, endorsed by 23 signatories, calling for:

• A global framework for free and open AI.
• Protection of freedom of expression, innovation, free trade, and access.
• Promotion of transparency and responsible use.
• Avoidance of government censorship or manipulation.
• Emphasis on international collaboration.

Conclusion

GTIPA's report underscores that AI policy must be proactive, flexible, and context-sensitive to drive innovation and global prosperity. It highlights the importance of regulatory stability, infrastructure investment, talent development, and international collaboration in shaping a successful AI ecosystem. The case studies and recommendations from member think tanks provide a roadmap for governments to foster AI growth while aligning with democratic values and market principles.