Operators are unwilling to be mere pipelines: China Mobile's ambitions and boundaries with AI-eSIM

Author | Huang Yu

Under the AI industrial revolution, telecom operators are destined to undergo a cross-era identity shift.

Recently, China Mobile announced it will launch the world’s first AI-eSIM product. Traditional SIM cards and eSIMs are essentially communication tools, limited to connectivity. The AI-eSIM, however, equips endpoint chips with a smart brain, a security foundation, computing power services, and communication connectivity—the four core capabilities.

For decades, operators served as “road builders,” transmitting traffic from base stations to phones. In the era of AI, China Mobile clearly doesn’t want to just be a pipeline collecting tolls. Through AI-eSIM, it is trying to become the “computing power grid” and “cloud brain” for smart hardware.

The underlying logic is clear: since the computing power of endpoints can’t keep up with the demands of large models, operators can directly deliver cloud computing power together with the network to devices.

Computing Power Business of Operators

In the perception of many, eSIM just integrates traditional SIM cards into chips, but China Mobile’s launch of AI-eSIM obviously goes much deeper.

Jiang Han, Senior Researcher at Pango Think Tank, told Wallstreet.cn that China Mobile’s pioneering launch of AI-eSIM is mainly to follow the trend of AI and IoT integration, meeting the market’s need for intelligent device connectivity and smart interactions.

According to China Mobile, the core highlight of this AI-eSIM is its ability to schedule cloud models in real-time, enabling devices to “think independently” and “respond instantly.”

Through cloud-network integration, China Mobile has deeply tied communication modules and AI computing power, granting end devices active thinking and lower interaction latency.

Feng Xiaodong, Head of Physical AI Product Line at Agora, pointed out to Wallstreet.cn that AI-eSIM enables devices to “be smart as soon as they're online,” marking operators’ transition from pipeline to intelligent computing power service providers.

In practical application, China Mobile has also identified initial key scenarios, including AI toys and smart wearables.

Jiang Han believes that as AI toys, smart wearables, and broad IoT devices grow rapidly, traditional SIM cards face limitations in remote management and flexible configuration. AI-eSIM can effectively solve issues such as instability, complex configuration, and insufficient intelligence, improving user experience and seizing opportunities in emerging smart hardware markets.

Take the currently hottest AI glasses, for example: AI glasses naturally have first-person sensing, but hardware is always plagued by the “impossible triangle” of weight, battery life, and performance.

Toubao Research Institute noted in the “2026 China Endpoint AI Industry Research Report” that, in the short term, AI glasses mainly serve as front-end interaction, responsible for voice wake-up, visual sensing, near-eye display, and scene-triggering. AI phones remain the computational base and ecosystem hub. In the mid-to-long term, with mature low-power SoC, Micro LED, waveguide optics, and independent connectivity, AI glasses may gradually evolve from an auxiliary entry to a more core platform.

AI-eSIM can solve AI glasses’ issues with independent connectivity, and also act as an “external brain” to make up for shortcomings in computing power and latency.

If equipped with AI-eSIM, when users issue commands to AI glasses, stability and responsiveness can be significantly improved.

Li Hongwei, CEO of Thunderbird Innovation, also told Wallstreet.cn that previously, smart glasses needed to pair with phones and connect via Bluetooth, but the user experience was suboptimal. Strategically, if smart glasses are to truly become the “next-generation phone,” they need to be independent—not always reliant on phones.

Clearly, AI-eSIM provides a critical solution for the next wave of experience upgrades for smart glasses, but the cellular connection enabled by AI-eSIM will undoubtedly put more pressure on battery consumption.

For a long time, operators’ headache has been “increased traffic but no increased revenue.” Although 5G user numbers keep rising, the fate of being a pipeline keeps operators at the bottom of the value chain. By contrast, terminal companies like Apple and Google reap most of the industry’s profits via their operating systems and app ecosystems.

China Mobile’s launch of AI-eSIM targets the pain points of “insufficient local computing power, unable to do high intelligence” in small edge devices—equipping small devices with a “shared cloud brain,” thus opening new growth spaces for its own business.

The Inevitable Path After Massive Investment

Behind any strategic action is the calculation of real money and the input-output ratio.

In recent years, operators have invested heavily in computing infrastructure. From "East Data, West Computing" nodes to nationwide edge data centers, China Mobile has accumulated massive computing resources.

In 2025, China Mobile will accelerate the layout of its computing power network, reaching a total scale of 92.5 EFLOPS (FP16), covering all computing specifications from hundreds to tens of thousands of cards.

At the same time, it has improved the three-tier computing power delay circle: city level 1 ms, provincial level 5 ms, national level 20 ms—its inter-provincial backbone 400G OTN network basically covers the country, with IDC standard racks exceeding 1.5 million.

Against the backdrop of plateaued traditional communication business growth, computing power services now form, along with communication and intelligent services, the three main businesses of China Mobile.

Chen Zhongyue, Chairman of China Mobile, recently stated publicly that China Mobile will strengthen full-stack innovation and integrated service capabilities for communication, computing, and intelligence, enhancing intelligent computing centers and cloud computing service supply.

In China Mobile’s 2025 performance report, it released its computing power service revenue for the first time, achieving 89.8 billion yuan, up 11.1% year-on-year.

The practical problem facing operators, including China Mobile, is how to absorb massive computing power and convert it into sustainable commercial income.

Jiang Han believes AI-eSIM is an important component of China Mobile’s AI strategy. It not only completes its intelligent connectivity service system but also facilitates computing power service output—empowering various smart terminals via cloud collaboration.

China Mobile’s launch of AI-eSIM is clearly a proactive move: rather than waiting for third-party apps to rent computing power, it’s better to bundle computing power directly into communication modules for sale.

When an AI-eSIM chip is embedded in an AI toy or AI glasses, China Mobile is no longer just selling data, call minutes, etc., but a subscription agreement that includes “connectivity + computing power + model service.”

Feng Xiaodong pointed out that AI-eSIM brings innovations to the AI toy and smart wearables hardware market, such as token accounts built into the AI-eSIM supporting billing by computing power consumption. China Mobile can bundle tokens for sale to customers—not only bringing incremental revenue, but also pushing AI hardware from one-off sales to a continuous AI subscription model.

This transformation is systemic. In the second half of 2025, China’s three major operators launched eSIM services; by 2026, the support has evolved from pure internet access to intelligent upgrades.

Viewed further, AI-eSIM is also China Mobile’s network occupancy for “next-generation smart terminals.”

Once smart glasses, watches, or even AI toys can run independently from phones, users’ dependency on communication terminals will shift. When hardware is no longer an accessory to phones, operators can use AI-eSIM to undertake this era change in hardware.

Future business models may be restructured as a result.

In this era where computing power equates to power, China Mobile’s AI-eSIM will undoubtedly bring a top-down rule restructuring to the hardware supply chain, redefining smart terminal connectivity, computing power allocation, and the commercial value landscape.

Risk Warning and DisclaimerThe market involves risks and investments require caution. This article does not constitute personal investment advice nor does it take into account individual users’ special investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions in this article suit their circumstances. Investment based on this is at your own risk.