2021-09-30-布鲁金斯学会-加强人工智能领域的国际合作进度报告（英）_123页_1mb

Summary of "Strengthening International Cooperation on AI"

Executive Summary

International cooperation in artificial intelligence (AI) is essential due to its collaborative nature, technical complexity, and potential societal impacts. Countries cannot effectively develop and deploy safe, trustworthy, and influential AI systems in isolation, as AI thrives on shared knowledge, resources, and standards. Without international cooperation, we risk divergent regulations, fragmented technology stacks, and barriers to AI's benefits.

This report argues that while complete harmonization of AI policies may not be feasible, significant convergence is not only desirable but necessary. The key lies in leveraging international platforms, standards development organizations, and joint research initiatives to foster collaboration without stifling innovation.

The recommendations include consolidating commitments to international cooperation in national policies, developing common definitions and principles for trustworthy AI, implementing risk-based approaches, enhancing transparency, building robust international standards, and funding participation in standards organizations. Launching joint large-scale R&D projects for global challenges represents another crucial avenue for cooperation.

Key Findings

1. AI is fundamentally collaborative: AI research, development, and deployment benefit immensely from international collaboration due to its resource-intensive nature, technical complexity, and global potential.

2. Regulatory divergence creates barriers: Divergent approaches to AI regulation increase compliance costs, limit market access, and fragment the global AI landscape.

3. Ethical alignment is crucial: Common principles for trustworthy AI development help focus global efforts and ensure ethical considerations are integrated across different jurisdictions.

4. Risk-based approaches provide flexibility: Moving beyond broad prohibitions or permissive stances toward nuanced risk classification allows for tailored regulatory responses.

5. International standards are essential: Harmonized standards facilitate interoperability, trust, and global adoption of AI technologies.

6. China's rising influence presents both opportunities and challenges: Engaging constructively with China while safeguarding democratic values remains a complex but necessary task.

7. Joint research projects catalyze innovation: Addressing global challenges like climate change and healthcare through collaborative AI research demonstrates tangible benefits of international cooperation.

Recommendations

1. Commit to international cooperation: Governments should formalize their commitment to collaborating on AI policies and development.

2. Develop common frameworks: Establish shared definitions of AI and ethical principles to guide international AI development.

3. Adopt risk-based regulation: Work toward harmonized approaches to risk assessment and management of AI systems across jurisdictions.

4. Enhance standards development: Support industry-led, research-driven international standards that are technically robust and globally applicable.

5. Fund participation in standards organizations: Increase financial support for academics and industry stakeholders in international standards bodies.

6. Launch joint R&D initiatives: Prioritize collaboration on AI projects addressing pressing global challenges like climate change and healthcare.

This comprehensive approach balances economic growth with ethical considerations, ensuring that the benefits of AI are widely distributed while mitigating associated risks through international collaboration.