Bloomberg: US AI pause fails without China transparency pact

Bloomberg Opinion columnist Catherine Thorbecke argues that unilateral U.S. pauses on artificial intelligence development will prove ineffective without reciprocal transparency commitments from China. Citing a simulated WeChat bug as illustrative evidence, Thorbecke contends that slowing the AI race requires verifiable global cooperation rather than isolated national restrictions. The column asserts that both Washington and Beijing must prioritize mutual disclosure of AI risks to prevent strategic miscalculations during competitive development cycles. This perspective challenges prevailing U.S. policy debates focused primarily on domestic safety regulations and export controls. Thorbecke suggests that without binding transparency mechanisms, American restraint merely cedes technological advantage while failing to mitigate existential risks. The analysis implies current diplomatic channels lack sufficient technical verification protocols for meaningful arms-control-style agreements. Consequently, the piece advocates for establishing joint risk-assessment standards before pursuing broader moratoriums. This position highlights the inherent tension between national security competition and collective safety governance in advanced AI development.