China's Digital Industry Reached 39.6 Trillion Yuan, but the Headline Hides a Measurement Problem

China reported 39.6 trillion yuan in digital industry revenue for 2025, an 8.8 percent increase that outpaced the country's overall economic growth.

The figure appears in the National Cyberspace Administration's newly released national informatization report. It arrived with another headline number: digital economy core industries contributed more than 10.5 percent of China's gross domestic product.

Those numbers describe an economy in which software, telecommunications, electronics, and digital services have become major growth engines. Yet they also create the central tension in the report. China has built digital infrastructure at extraordinary scale, but measuring the economic return from that infrastructure remains surprisingly difficult.

A separate official estimate published months earlier placed 2025 digital industry revenue at approximately 38.3 trillion yuan. That 1.3 trillion yuan difference does not automatically make either number wrong. It does show why readers must distinguish revenue, value added, consumption, infrastructure coverage, and productivity.

The report offers strong evidence that China's digital capacity expanded in 2025. It provides less clarity about how evenly that capacity translated into business returns, household welfare, and sustainable productivity gains.

The 39.6 Trillion Yuan Headline Comes With a Wider Scorecard

The report's most important message is not one revenue figure. It is the simultaneous expansion of digital production, consumption, and network capacity.

The Cyberspace Administration of China released the informatization report on July 24, 2026. The document reviews progress made during 2025, the final year of China's 14th Five-Year Plan.

It says revenue from the country's digital industries reached 39.6 trillion yuan, rising 8.8 percent from the previous year. Digital economy core industries produced value added equal to more than 10.5 percent of national GDP.

These measures describe related but different parts of the economy. Revenue is the total income recorded by businesses before costs and intermediate purchases are removed. Value added measures the additional economic value created during production.

That distinction matters because a large revenue base does not translate directly into an equally large contribution to GDP. Hardware supply chains, cloud services, telecommunications, and online platforms often buy substantial inputs from other businesses.

The report also places China's digital consumption at 25.3 trillion yuan, up 8.7 percent. Digital consumption can include online goods, digitally delivered services, connected devices, and spending enabled by digital platforms.

The exact boundaries depend on the statistical framework. However, the figure suggests that the digital economy is no longer confined to technology companies. It now shapes retail, entertainment, transportation, finance, education, and household services.

China's total GDP reached 140.1879 trillion yuan in 2025, according to the country's economic communiqué. GDP increased 5 percent during the year.

Digital industry revenue therefore grew 3.8 percentage points faster than GDP. The comparison is imperfect because revenue and GDP use different accounting concepts. Still, it indicates that digital businesses expanded faster than the broader economy.

The sector's composition also matters. Software and information technology services generated 15.4831 trillion yuan in revenue, an increase of 13.2 percent. That growth rate exceeded both overall GDP growth and the reported 8.8 percent increase in digital industry revenue.

Information transmission, software, and IT services produced 7.0599 trillion yuan in value added. Their real growth rate reached 11.1 percent, making the category one of the faster-growing parts of the national economy.

The report therefore describes more than infrastructure construction. Software, cloud computing, data services, and digital platforms increasingly account for the growth layered on top of China's networks.

That shift creates a tougher benchmark for the next planning period. Adding connections remains important, but the harder question concerns what businesses and households accomplish with them.

Infrastructure Has Moved From Scarcity to Saturation

China's network story is shifting from whether access exists to whether organizations can use that access productively.

China ended 2025 with 4.838 million 5G base stations. That total represented a net increase of 588,000 during the year, according to the national telecommunications report.

The country had 34.4 5G base stations for every 10,000 residents. That level exceeded the relevant national planning target by 8.4 stations.

Every county had access to gigabit optical networks. The government also reported 5G coverage in every town and more than 95 percent of administrative villages.

Two-thirds of prefecture-level cities met China's gigabit-city standard. Such standards evaluate more than advertised broadband speeds. They can include network availability, user adoption, service quality, and industrial applications.

By the end of 2025, 1.204 billion mobile subscriptions used 5G services. China also had 238.39 million fixed broadband accounts with speeds of at least one gigabit per second.

IPv6 active users reached 869 million. IPv6 is the newer internet addressing protocol designed to support far more connected devices than the older IPv4 system.

These figures show why simple coverage goals are losing explanatory value. A network reaching nearly every populated area no longer guarantees that each additional base station produces the same benefit.

Early network construction can eliminate a basic access barrier. Later investment must improve capacity, reliability, latency, energy efficiency, or coverage inside factories and buildings.

That transition is visible in the rise of 5G-Advanced, an upgraded stage of 5G that supports higher capacity and more precise industrial connections. By the end of 2025, 5G-Advanced service covered more than 330 Chinese cities.

China's three major telecommunications operators also offered 938,000 data center racks. Their investment focus is moving from broad resource deployment toward coordinated computing and network services.

The distinction affects companies buying infrastructure. A manufacturer does not gain productivity merely because a nearby tower supports 5G. It needs compatible equipment, usable software, process redesign, trained employees, and a business case that justifies deployment.

China reported more than 23,000 projects combining 5G with the industrial internet. The industrial internet connects machinery, sensors, production software, and business systems across manufacturing operations.

Those projects span all 41 major industrial categories. However, a project count does not reveal whether applications remain pilots or become repeatable operating systems across entire companies.

The same caution applies to consumers. Gigabit broadband availability does not show whether households need that capacity or receive meaningful new services from it.

Infrastructure saturation changes what good performance looks like. Policymakers must increasingly evaluate utilization, service quality, operational savings, and economic output rather than counting installed equipment.

Software Growth Is Becoming More Important Than Tower Growth

The fastest digital growth came from software and services, suggesting that value is moving upward from physical networks into applications.

China's software industry generated 15.4831 trillion yuan in 2025, according to official software statistics. Revenue increased 13.2 percent, while total profit rose 7.3 percent.

The difference between revenue and profit growth deserves attention. It suggests that the sector expanded quickly, but companies did not convert all that expansion into proportional earnings.

Several factors can create that pattern. Businesses might invest heavily in artificial intelligence infrastructure, compete through lower margins, or depend on labor-intensive implementation work.

Information technology services accounted for 68.7 percent of software industry revenue. This category grew 14.7 percent and included cloud computing, big data services, systems integration, and other technical work.

Cloud computing and big data services generated 1.623 trillion yuan, increasing 13.6 percent. Integrated circuit design revenue reached 442.1 billion yuan after growing 18.9 percent.

Those categories connect the infrastructure story to the application economy. Data centers create useful capacity only when software companies, enterprises, and public agencies turn computing resources into deployable services.

The transition is especially visible in artificial intelligence. AI phones, personal computers, glasses, and service robots moved further into commercial distribution during 2025.

China produced 18.581 million service robots, an increase of 16.1 percent. Production data shows manufacturing activity, although it does not reveal how many products reached sustained use.

Software also creates an important regional contrast. China's eastern provinces produced 83.1 percent of national software revenue. The Yangtze River Delta alone accounted for 28.9 percent.

Beijing, Guangdong, Jiangsu, Shandong, and Shanghai led provincial software revenue. This concentration reflects existing clusters of technical talent, research institutions, customers, financing, and major technology companies.

Digital infrastructure may be geographically broad, but the economic value built on that infrastructure remains concentrated. A western city can receive advanced connectivity without immediately developing the same software ecosystem as Beijing or Shanghai.

This is the main pressure point for regional policy. Governments can fund networks, computing centers, and industrial parks. They cannot quickly manufacture dense markets for skilled workers, enterprise customers, or experienced software founders.

Businesses face a related challenge. Buying cloud capacity or AI tools is easier than redesigning workflows around them.

A factory must connect operational technology with enterprise software while protecting safety and production continuity. A hospital must integrate data across systems without weakening patient privacy. A retailer must decide whether automation improves conversion, logistics, or customer support.

For knowledge workers, the value often depends on whether information can move across meetings, files, notes, and applications. A searchable knowledge base can reduce retrieval work, but only when organizations maintain reliable source material and access controls.

These examples show why software revenue is a better signal than tower counts for the next phase. Applications determine whether network capacity becomes lower costs, better products, or merely another operating expense.

The Numbers Reveal a Measurement Conflict

China's digital economy is large by every available measure, but the reported totals do not yet form one transparent accounting picture.

The new report places 2025 digital industry revenue at 39.6 trillion yuan. An earlier government account described the total as approximately 38.3 trillion yuan.

The earlier figure appeared in January 2026 and was attributed to the Ministry of Industry and Information Technology. It also placed industry profit at 3.1 trillion yuan.

A difference of 1.3 trillion yuan is about 3.4 percent of the lower estimate. That is large enough to matter, even within an economy of China's scale.

There are several plausible explanations. The July report may use revised data, broader industry boundaries, more complete submissions, or a different treatment of overlapping activities.

The January number was explicitly described as an initial calculation. Preliminary economic data commonly changes when agencies receive additional records or reconcile different reporting systems.

The agencies may also be measuring different statistical populations. "Digital industry" can cover telecommunications, software, electronics manufacturing, internet services, data services, and selected emerging technologies.

Small changes in category boundaries can move large amounts of revenue. Electronics manufacturing alone contains extensive supply chains with substantial intermediate sales.

Neither figure should be treated as false without a methodological reconciliation. However, publishers and analysts should not repeat 39.6 trillion yuan as if it were a self-explanatory measure.

The same problem affects the reported 10.5 percent share of GDP. Digital economy core industries have a formal statistical definition, but the broader digital economy extends into nearly every conventional sector.

A digitally managed factory still produces manufacturing output. An online retailer remains part of wholesale and retail trade. A bank using AI still produces financial services.

Analysts can count the direct value added of technology industries with reasonable consistency. Measuring digitally enabled value across traditional industries involves more assumptions.

Digital consumption creates another boundary problem. A phone purchased online is a physical product, a digital device, and an e-commerce transaction. Statistical frameworks must decide which attributes qualify it for inclusion.

This does not make digital economy estimates useless. It means that trend consistency is often more informative than one absolute total.

Readers should ask whether agencies use the same definition across years. They should also examine whether reported growth comes from prices, production volumes, category changes, or newly included businesses.

Cross-country comparisons require even more caution. The United States, European Union, and China do not always define core digital activities identically.

A country with a large electronics manufacturing base can report more digital industry revenue than one that imports devices. Another economy may generate more value from software and intellectual property despite recording lower supply-chain revenue.

Revenue can also count the same economic chain at several stages. A component supplier sells to a device maker, which sells to a distributor, which sells to a consumer.

GDP accounting removes intermediate consumption to reduce such duplication. This is why the report's value-added share and industry revenue should never be presented as interchangeable measures.

The strongest conclusion survives the discrepancy. China's digital sectors grew faster than the national economy during 2025, led by software and IT services.

The weaker conclusion is that 39.6 trillion yuan precisely captures the sector's economic footprint. The published materials do not provide enough methodological detail to support that interpretation.

Scale Does Not Guarantee Equal Returns

The central policy challenge is no longer building the world's largest digital system. It is extracting durable returns without widening regional and organizational gaps.

China's digital network gives domestic companies an unusually large environment for testing connected products. More than one billion 5G subscriptions can support services that require broad mobile availability.

Industrial companies can deploy connected machinery across factories and supplier networks. Consumer platforms can distribute software updates, AI features, and digital payments at national scale.

This scale reduces one constraint, but it exposes several others. Companies still need proprietary data, experienced employees, integration budgets, cybersecurity controls, and customers willing to pay for new services.

Large businesses generally possess more of those resources than smaller firms. They can run pilots across several departments, absorb failures, and hire specialist teams.

Small businesses may depend on standard cloud products or outside vendors. That can speed adoption, but it can also leave them with fragmented systems and limited control over data.

Regional inequality presents a similar problem. Nationwide coverage can narrow the access gap while software revenue remains clustered in major coastal economies.

The ten leading provincial regions generated more than 90 percent of the country's digital industry revenue growth, according to the earlier official account. That concentration shows that connectivity alone does not distribute commercial capability.

A city seeking software growth needs more than a data center. It needs customers with complex technology needs, universities producing skilled workers, credible career paths, and companies able to retain experienced managers.

Infrastructure investment can even produce weak returns when local demand is overstated. Underused data centers consume capital and energy without creating a corresponding base of applications.

Artificial intelligence raises the stakes. Training and serving AI systems require computing capacity, but utilization can vary significantly across facilities.

A rack count does not reveal whether servers operate near capacity, support high-value workloads, or remain tied to promotional projects. Energy efficiency and workload quality matter alongside physical installation.

Cybersecurity creates another cost. More connected factories, vehicles, hospitals, and public systems expand the number of potential entry points for attackers.

IPv6 adoption supports a larger device ecosystem, but it does not automatically improve security. Organizations still need correct configuration, identity management, monitoring, and incident response.

Data governance will influence whether businesses can combine information across systems. Restrictive or unclear rules can slow useful integration, while weak controls can expose personal or commercially sensitive information.

Consumers face their own tradeoff. Digital services improve convenience, yet extensive data collection can increase surveillance, profiling, fraud, and platform dependence.

The national report says surveyed internet users associated digital technologies, including AI, with wider learning access, improved work efficiency, and better quality of life. That finding represents reported sentiment, not a controlled measure of productivity.

Surveys can capture whether users notice benefits. They cannot isolate whether those benefits outweigh costs such as distraction, privacy loss, or employment disruption.

China's 25.3 trillion yuan in digital consumption therefore needs supporting indicators. Household adoption, satisfaction, service quality, fraud losses, and time savings would show more about welfare than spending alone.

A similar standard applies to enterprise adoption. Companies should report whether digital systems reduce downtime, increase output, shorten development cycles, or improve inventory accuracy.

Workers also need a practical way to manage the growing volume of digital material. A personal second brain can help organize scattered information, but tools cannot correct unreliable data or poorly designed work.

The distinction between access and capability will define the next stage. China has largely solved basic digital availability. The harder work concerns organizational adoption and measurable results.

What to Watch After China's Digital Industry Report

Three signals will show whether China's digital scale is becoming a stronger economic engine or simply a larger installed base.

The first signal is statistical reconciliation. Future releases should explain why the 39.6 trillion yuan estimate differs from the earlier 38.3 trillion yuan figure.

A clear bridge between preliminary and revised totals would strengthen confidence in the headline. It should identify category changes, data revisions, and any expanded reporting population.

The most useful disclosure would separate telecommunications, electronics manufacturing, software, internet services, and other digital activities. It would also show revenue, profit, and value added for each category.

Without that detail, analysts can observe growth but cannot confidently identify its source. A rise driven by low-margin hardware sales has different implications from one driven by software exports or cloud services.

The second signal is software profitability and business composition. Software revenue grew 13.2 percent, while profit increased 7.3 percent.

If profit growth moves closer to revenue growth, the sector may be converting expansion into healthier operating returns. If the gap widens, competition and implementation costs may be consuming more of the gains.

Cloud, big data, information security, industrial software, and integrated circuit design deserve separate attention. Their growth rates will show where companies are finding sustainable demand.

Software exports offer another useful test. China reported 62.73 billion dollars in software exports for 2025, up 7.7 percent.

Faster export growth would suggest that Chinese software companies are earning more revenue beyond the domestic market. Slower growth would reinforce the view that the sector still depends heavily on internal demand.

The third signal is infrastructure utilization. Official updates should move beyond base station, rack, and subscriber totals.

For 5G, useful measures include traffic per site, enterprise adoption, service reliability, and the share of industrial pilots that reach regular production. For data centers, utilization rates and energy use per computing workload would provide more economic meaning.

For gigabit networks, policymakers should examine how many subscribers actually purchase high-speed connections and which applications require that capacity. Availability and adoption are different achievements.

Industrial internet projects need outcome measures such as reduced downtime, higher output, lower defect rates, and faster maintenance. A project count alone cannot establish return on investment.

These signals will also show who faces pressure. Telecommunications operators must justify continued investment after reaching broad national coverage.

Cloud providers must turn installed computing capacity into recurring workloads. Software vendors must deliver measurable customer outcomes while protecting margins.

Regional governments must demonstrate that digital infrastructure attracts productive activity rather than duplicating facilities. Enterprises must move successful pilots into daily operations.

The new report establishes a credible scale story. Digital industry revenue grew faster than China's economy, software expanded even faster, and network access reached extraordinary breadth.

Its harder message is that scale has stopped being sufficient evidence of progress. The next national informatization scorecard will need to explain not only what China built, but what that system produces.

Watch the revised industry accounts, software margins, and infrastructure utilization over the next year. Together, they will reveal whether China's 39.6 trillion yuan digital sector is becoming more productive, or merely becoming larger.