ZTE’s chief development officer Cui Li outlined a new AI‑centric roadmap at the Mobile World Congress, stressing a shift from static models to adaptable, data‑driven systems.
ZTE Corporation (HKEX: 0763, SZSE: 000063) used its keynote slot at MWC Shanghai 2026 to lay out a comprehensive AI strategy that moves the Chinese telecom giant beyond incremental upgrades toward a fundamentally data‑centric operating model. The address, titled “Unlocking Value and Embracing Uncertainty in the AI Era,” was delivered by Cui Li, ZTE’s Chief Development Officer, on June 25, 2026.
Core Messages from the Shanghai Stage
Cui Li framed the current AI landscape as one where “uncertainty is the only certainty.” He argued that the rapid pace of model iteration and the rise of highly customized workloads render the traditional “one‑size‑fits‑all” approach obsolete. In response, ZTE is pursuing an “AI‑centric” strategy built around four guiding pillars:
- Openness and Decoupling – Designing components that can be swapped or upgraded without disrupting the broader system.
- Flexible Scaling – Enabling workloads to expand or contract in real time, matching the unpredictable demand spikes typical of generative AI services.
- Extreme Synergy – Tight integration between compute, storage, and networking layers to minimize latency and maximize throughput.
- Scenarios First – Prioritizing real‑world use cases—such as edge analytics, autonomous networking, and AI‑driven customer experiences—over abstract benchmark scores.
These points were presented as a response to the “paradigm shift” the speaker sees sweeping across the industry, where AI is no longer a peripheral add‑on but a core engine for product and service innovation.
The “All‑in‑AI, AI for All” Playbook
ZTE’s roadmap is anchored in two parallel objectives:
- Embedding AI‑native capabilities across the portfolio. By weaving intelligence into hardware, software, and services, ZTE aims to deliver a “great leap in value delivery” for enterprise customers.
- Transitioning to a data‑driven organization. The company is restructuring internal processes to foster human‑machine collaboration, allowing engineers to leverage AI assistants for design, testing, and operations.
Cui Li emphasized that this dual thrust is not merely a marketing tagline but a concrete shift in architecture. The firm plans to replace monolithic, static pipelines with modular, AI‑augmented workflows that can be reconfigured on the fly.
Building a Resilient AI System
Resilience, according to ZTE, is a function of agility and rapid evolution. The four dimensions outlined above serve as a blueprint for constructing systems that can:
- Adapt to new model architectures without extensive re‑engineering.
- Scale compute resources dynamically, supporting bursty generative workloads while keeping cost under control.
- Maintain high performance through tight coupling of hardware accelerators and software stacks.
- Deliver consistent outcomes across diverse deployment scenarios—from 5G edge nodes to cloud‑scale data centers.
This approach mirrors broader industry trends where vendors are moving away from static AI stacks toward elastic, service‑oriented AI platforms that can be quickly repurposed for emerging use cases.
Why Enterprises Should Take Note
For B2B customers, the implications are threefold:
- Reduced Vendor Lock‑In – Openness and decoupling mean that enterprises can integrate best‑of‑breed components without being forced into a single‑vendor ecosystem.
- Faster Time‑to‑Value – Flexible scaling and scenario‑first design promise quicker deployment of AI‑driven services, a critical factor in competitive markets.
- Enhanced Operational Efficiency – Extreme synergy reduces data movement overhead, translating into lower latency and power consumption—key metrics for edge deployments and large‑scale cloud operations.
Cui Li’s remarks also hint at a future where customer experiences become a competitive differentiator. By embedding AI assistants into development and operational workflows, ZTE is betting that enterprises that adopt such collaborative models will outpace rivals still relying on manual processes.
Industry Context
ZTE’s announcement arrives at a moment when generative AI, large language models (LLMs), and AI‑driven networking dominate vendor roadmaps. Competitors such as Huawei, Nokia, and Ericsson have similarly emphasized AI‑centric architectures, but ZTE’s explicit focus on resilience and uncertainty management sets it apart.
The “All‑in‑AI” narrative also aligns with the broader MLOps movement, which stresses reproducibility, continuous integration, and automated monitoring of AI pipelines. By framing its strategy around openness, scaling, and scenario orientation, ZTE is positioning itself to support the full lifecycle of AI—from model training to edge inference.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
