Article illustration 1

Three days after the Trump administration released its America-first AI strategy, China countered with a meticulously timed geopolitical maneuver. At the opening of Shanghai's World Artificial Intelligence Conference (WAIC), Premier Li Qiang unveiled Beijing's "Global AI Governance Action Plan"—a blueprint positioning China as the standard-bearer for international AI cooperation amid conspicuous American absence.

The Shanghai Consensus

Against a backdrop of Western no-shows (only Elon Musk's xAI represented U.S. labs), WAIC became an unexpected epicenter for AI safety discourse. "You could literally attend AI safety events nonstop," observed Brian Tse of Concordia AI, which hosted a safety forum featuring Turing Award winners Yoshua Bengio and Stuart Russell. Chinese researchers like Shanghai AI Lab’s Zhou Bowen showcased advanced safety monitoring frameworks, while Professor Yi Zeng of the Chinese Academy of Sciences advocated for multinational coalitions: "It would be best if the UK, US, China, Singapore institutes come together."

The Great Role Reversal

The geopolitical irony is palpable. Where China was once seen as constrained by censorship, its new plan champions UN-led governance and proactive regulation. Meanwhile, the Trump administration’s push for AI models to "pursue objective truth" resembles the ideological control it once criticized. "We’ve witnessed a complete role reversal," notes Paul Triolo of Albright Stonebridge Group. "With the U.S. absent, a coalition co-led by China, Singapore, the UK, and the EU is building guardrails for frontier AI."

Convergent Risks, Divergent Policies

Technical realities bind the AI superpowers: hallucinations, bias, and cybersecurity threats manifest similarly across borders due to shared model architectures. Yet policy responses diverge radically. While China implements binding domestic standards and Xi Jinping personally advocates guardrails, the Trump administration recently attempted to block state-level AI regulations. This governance vacuum, experts warn, cedes soft power to Beijing at a pivotal moment.

Industry’s Hesitant Embrace

Despite governmental enthusiasm, Chinese tech giants mirror Western counterparts’ reluctance. Concordia AI’s analysis reveals only 3 of 13 major Chinese AI developers publish safety assessments. WAIC entrepreneurs privately admitted prioritizing scaling and monetization over existential risks. But pressure is mounting—researcher Bo Peng argues China’s open-source push (like his RWKV model) could force transparency: "Different AIs embodying different values will keep each other in check."

As Triolo predicts increased Chinese safety publications, Beijing’s calculus becomes clear: In the void left by American retreat, the race for AI supremacy now includes a parallel contest to define its ethical boundaries.