Global AI governance needs multilateral, not unilateral, rules: why China’s push reveals power gaps and cultural blind spots
Original framing: “Should China aim for the lead in making AI rules for the world?” — South China Morning Post
Peer‑reviewed AI safety research highlights risks of model scaling, emergent capabilities, and bias amplification, underscoring the need for transparent benchmarking and reproducibility standards. Scientific bodies such as the International Association for AI Safety advocate for open‑source evaluation suites, which remain underutilised in policy drafts. Robust scientific input can anchor governance in evidence rather than geopolitical posturing.
The rush to position China as the de‑facto AI regulator masks a deeper governance vacuum that stems from historic failures to collectively manage transformative technologies.
By integrating Indigenous data sovereignty, cross‑cultural ethical frameworks, and rigorous scientific benchmarking, a multilateral treaty can prevent the concentration of power and ensure that AI serves global public goods. Capacity‑building in the Global South and transparent ethics councils will democratise innovation, while a community‑stewarded data commons safeguards cultural integrity. Together, these mechanisms echo past successes in nuclear and internet governance, offering a resilient, inclusive pathway for AI’s future.