Summary of Key Points
China has released its first comprehensive document focusing on international cooperation in artificial intelligence (AI), titled the "Action Plan for Cooperative Development in Artificial Intelligence." Addressing current global challenges in AI development, such as the disparity in computing power, data barriers, talent shortages, and fragmented governance, the plan outlines specific actions across eight areas: data, computing power, open source, empowerment, talent, governance, ethics, and transparency. The aim is to promote AI as a global public good, fostering collaborative governance and inclusivity, so that the benefits of technology can benefit all humanity.
Why This Action Plan?
AI is developing rapidly, but it also faces numerous challenges:
- Uneven Resource Distribution: Some countries have ample computing power and data, while others lack even basic AI tools (for example, small African nations may struggle to use AI for agriculture due to insufficient resources).
- Controllable Technical Risks: Issues like algorithmic biases (discrimination against women in recruitment), data breaches (AI stealing user information), and deepfakes (fake videos) cannot be effectively managed by a single country.
- Fragmented Governance: Different countries have different regulations; for instance, China's AI security laws are not compatible with those of the European Union, making it difficult for companies to conduct cross-border AI projects.
Therefore, while the question of whether to cooperate was once optional, how to cooperate has become imperative—this is the context behind the introduction of the plan.
What Does the Action Plan Aim to Achieve?
The plan covers the entire spectrum of AI development, from "basic resources" to "advanced applications," and from "technology" to "ethics":
1. Data: Ensuring data is secure and accessible. The plan aims to create a "credible data space" where data can be shared securely between countries (for example, medical AI projects that require global patient data can now proceed with confidence).
2. Computing Power: Making AI affordable for all countries. The plan promotes the interconnection of computing facilities worldwide and provides affordable computing services to developing nations (for instance, subsidizing small countries to use AI for agricultural forecasting).
3. Open Source: Promoting the sharing of tools and algorithms. It aims to establish international open-source communities and set rules to prevent the misuse of open-source technology (e.g., preventing the use of AI in creating fake news).
4. Empowerment: Using AI to support various industries (e.g., aiding developing countries with agriculture, healthcare, and education through AI). Cross-border cooperation platforms will be established for collaborative projects.
5. Talent Development: Jointly training AI professionals, establishing global standards for certifications, and improving public awareness of AI (e.g., teaching the elderly how to use AI for medical purposes), while also addressing employment concerns.
6. Governance: Developing unified rules to address the issue of fragmented governance. This includes setting global AI safety standards and ensuring that national regulations are compatible with each other.
7. Security: Establishing mechanisms to prevent the misuse of AI, such as sharing information on cyber threats and making AI decisions more transparent.
8. Ethical Use: Ensuring that AI is used in a humane manner, eliminating biases and promoting its positive impact on society (e.g., helping vulnerable groups).
The Global Implications of This Plan
- Turning Concepts into Action: It transforms the idea of AI as a public good from a slogan into a reality (for example, making AI accessible to all through affordable computing power).
- Filling Institutional Gaps: The plan fills existing gaps in global AI governance by proposing new frameworks for cross-border data sharing and equitable access to resources.
- Supporting Developing Countries: It prioritizes the needs of developing nations, providing them with necessary resources and support to avoid being left behind in the AI era.
- A New Model of Governance: It avoids a one-size-fits-all approach and seeks a balance that respects cultural differences while finding common ground.
How to Implement the Plan
- Establishing Sustainable Mechanisms: Creating specialized organizations to promote project implementation and establishing legal and technical standards for secure data sharing.
- Involving Multiple Stakeholders: Government agencies, companies (such as Huawei and Google), educational institutions (like Tsinghua University and MIT), and non-governmental organizations (NGOs) should all be involved in implementing the plan.
- Adaptive Approach: The plan is designed to evolve with rapid technological advancements by conducting annual evaluations and updating relevant regulations.
In summary, this plan is not just a Chinese initiative; it aims to leverage global cooperation to ensure that AI benefits everyone, rather than being a privilege of a few countries. While its implementation requires the joint efforts of all nations, the direction is clearly positive.