Conversation with a ZTE (中兴) executive: As Silicon Valley is mired in large‑model infighting, how can China’s AI accelerate productivity monetization for government and enterprise?

The pitch: practical monetization over model warfare

A senior executive from ZTE (中兴) told reporters it has been reported that China’s AI push should pivot from headline‑grabbing large models to concrete productivity gains for government and enterprise customers. The executive framed the current moment as one where Silicon Valley is “mired in large‑model infighting” — debates over model scale, open‑sourcing and safety — while Chinese vendors can capture value by embedding AI into telecom networks, manufacturing lines and public services. The central claim: pragmatic, measurable ROI will win contracts and scale faster than chasing model size alone.

What that means for products and customers

For Western readers unfamiliar with China’s landscape, the shift is significant. Chinese tech firms often sell directly to government agencies and state‑owned enterprises, where procurement favors reliability, data sovereignty and predictable cost savings. It has been reported that ZTE is focusing on edge deployments, vertical domain models and integrated service contracts — not just API access to a general‑purpose LLM — so that efficiency improvements translate into recurring revenue for both vendors and clients.

Geopolitics and industrial policy as accelerants

This strategy is shaped by geopolitics. ZTE itself has a history with U.S. trade restrictions, and broader export controls on chips and software have reportedly pushed Chinese suppliers to double down on domestic ecosystems and government partnerships. Beijing’s industrial policy and procurement levers make the government both a customer and a regulator; together they can fast‑track pilots into city‑ or industry‑wide deployments in ways that differ from market‑led adoption in the West.

From pilots to payment: the remaining gaps

Practical hurdles remain: standardized metrics for productivity, data sharing agreements, and aligning operational workflows so AI recommendations are actually adopted. How do you convert efficiency gains into billable services? The ZTE executive argued for outcome‑based contracts, verticalized models, and clearer compliance frameworks to bridge pilots and large‑scale monetization. If China’s AI ecosystem can solve that commercial puzzle, the result may be less about who trains the biggest model and more about who reliably turns intelligence into income.