AI and Technical Standardization in China and the EU: Diverging priorities and the need for common ground

Given the highly disruptive potential of AI, global cooperation on AI safety and governance is imperative, and yet the deeply transformational potential of AI also ensures that a high level of competition and systemic rivalry is likely unavoidable. How can the EU best manage its complex relationship with China in the field of AI so as to ensure a necessary level of cooperation in spite of competition and rivalry? This Policy Brief offers insights from the field of technical standardization for AI. Technical standards are crucial for defining the parameters of AI systems, from basic reference architectures to security and ethics requirements to the technical functioning of specific applications in a wide diversity of fields including healthcare, education, advanced manufacturing, energy, and agriculture. In their efforts to harness and channel the development of AI, both China and the EU have turned to technical standard setting as a means to mitigate risks and achieve broad political goals. The EU’s AI Act has placed technical standards at the heart of the AI governance conversation by aiming for the development of European “harmonized standards” around risk criteria that AI systems and products must meet in order to comply with EU regulations. China meanwhile has aimed to establish itself as a global leader in AI standardization and is working to balance two, often competing priorities of ensuring control while facilitating innovation and boosting competitiveness. The EU and China seem to be at odds, and yet, common standards are needed to ensure space at the foundational, technical level for necessary cooperation on AI safety and governance and to avoid a more structural slide into de-coupling.