January 9 news.Central Radio and TelevisionChina Institute for the Development of the Electronic Information Industry of the Ministry of Joint Industrial and Technological Affairs, the Science City Management Committee of Sino-Kangmura, the Management Committee of the New Technologies Development Zone of Wuhandong Lake, the Chinese University of Science and Technology, the China-China University of Science and Technology, and the Shanghai Comprehensive National Science CentreAIResearched by institutions like the Institute of Artificial Intelligence and Big Data Research Institute and the Science and Technology Institute of Chinatrend.
1AI is summarized below:
AI GOVERNANCE GLOBALIZATION
Artificial Intelligence for All is at the heart of the global development agenda。
China has initiated the establishment of the World Organization for Artificial Intelligence Cooperation and hopes to actively provide the international community with artificial intelligence public goods through cooperation in development strategies, governance rules, technical standards, etc. Technology and infrastructure development, which seeks to bridge the digital divide and deepen international cooperation in artificial intelligence development and governance, are important in promoting the healthy development of artificial intelligence, promoting world economic growth and addressing global challenges such as climate change, public health and education equity。
Smart calculator size
SUPPLY OF KEY INDUSTRIAL ELEMENTS WILL BE FURTHER ENHANCED. NATIONALLY PRODUCED AI CHIPS WILL BE SCALED UP IN SPECIFIC SCENARIOS, AND THE COMPUTING INFRASTRUCTURE THAT SUPPORTS THE DEVELOPMENT OF LARGE MODELS WILL BE IMPROVED. THE RAPID DEVELOPMENT OF NATIONAL PRODUCTION CHIPS, THE DEVELOPMENT OF NEW STRUCTURES SUCH AS DEDICATED INTEGRATED CIRCUITS (ASIC) AND INTEGRATED STORAGE SYSTEMS WILL FACILITATE TECHNOLOGY OUTBURSTS AND THE GRADUAL DEVELOPMENT OF CO-ECOSYSTEMS FOR SOFTWARE AND HARDWARE. THE VANCA-CLASS CLUSTER BECAME THE MAIN VEHICLE SUPPORTING LARGE-SCALE MODEL TRAINING, WITH BREAKTHROUGHS IN SUPER-LARGE-SCALE CLUSTER TECHNOLOGY, AND HIGH-SPEED INTERCONNECTIONS SYNCHRONIZED WITH LOW-CARBON GREEN TECHNOLOGIES. THE EAST AND WEST COUNT PROJECT PROMOTES THE COORDINATED DEPLOYMENT OF NATIONAL COMPUTING RESOURCES AND SIGNIFICANTLY ENHANCES THEIR INCLUSIVE NATURE。
Application mainstreaming
AI SMART BODY FULLY ENTERS THE SCENE. IN 2026, AI ' S APPLICATION OF THE DEVELOPMENT PARADIGM WILL SHIFT FROM THE PURSUIT OF UNIVERSAL CAPABILITIES TO A DEEP-SEATED SOLUTION OF THE OCCUPATIONAL PAINS IN VERTICAL AREAS. AS THE TECHNOLOGY, SUCH AS THE INTELLIGENCE CLUSTER AND THE INTERACTIVE PROTOCOLS, MATURES, ENTERPRISE-LEVEL SMARTS WILL BE SCALED UP AT CORE BUSINESS SEGMENTS, SUCH AS R & D, PASSENGER SERVICE, OFFICE AUTOMATION, AND WILL GRADUALLY HAVE THE CAPACITY TO PROCESS FULL BUSINESS CLOSED LOOPS, MARKING THE ACCELERATION OF AI INNOVATION FROM LABORATORY TO REAL PRODUCTIVITY. AT THE POLICY LEVEL, THE OPINION ON THE IMPLEMENTATION OF THE "ADVISORY INTELLIGENCE + MANUFACTURING" INITIATIVE PROVIDES THAT BY 2027, 1,000 HIGH-LEVEL INDUSTRIAL INTELLIGENCE BODIES WILL HAVE BEEN LAUNCHED, CREATING AN ENABLING APPLICATION SERVICE FOR "INTELLECTUAL, WELL-ESTABLISHED" INDUSTRIES, WITH THE AIM OF NURTURING INDUSTRIAL INTELLIGENCE AND INTELLECTUAL PROGENY. THE STATE COUNCIL'S OPINION ON THE IN-DEPTH IMPLEMENTATION OF ARTIFICIAL INTELLIGENCE PLUS STATES THAT BY 2030 THE USE OF SMARTS WILL HAVE EXCEEDED 90%。
Multimodular Operationalization
Artificial intelligence core technologies move from “special tools” to “general intelligence partners”。
In 2025, large models such as DeepSeek brought “high-performance, low-cost” breakthroughs, driving the threshold and cost of artificially intelligent technology applications down significantly, opening new paths to global large model development. Currently, global algorithm upgrades support the long context of million-degree Token, which promotes multi-modular interactions between humans and “what they see” by integrating multi-source data such as text, images, audio, video and 3D dot clouds. At the same time, the “world model” with the ability to reason and plan is becoming the focus of competition and, by simulating the consequences of action, is driving AI from perception to decision-making intelligence。
UNIVERSALIZATION OF PRIMARY AI TERMINAL HARDWARE
The new generation of smart terminals is integrated with the immersion experience consumption scenario。
IN 2025, THE CONSUMER ELECTRONICS SECTOR SHOWED A MARKED TREND TOWARDS DIFFERENTIATION. ON THE ONE HAND, THE INTEGRATOR OF HARDWARE PARAMETERS IS INCREASINGLY CLOSE TO THE PHYSICAL AND COST CAPS, AND THE CONVERGENCE OF TECHNOLOGIES, SUCH AS FOLDING SCREENS, VIDEO SENSORS, AND FAST-FILLING, IS BECOMING MORE HOMOGENEOUS AND MARKETS ARE COMPETING IN THE RED SEA; ON THE OTHER HAND, AI MOBILE PHONES AND VARIOUS TYPES OF AI HARDWARE HAVE CONTINUED TO GROW AND HAVE BEEN PROMINENT IN TERMS OF VOLUME, USER ATTENTION AND ECOLOGICAL CONSTRUCTION, AS KEY DRIVERS OF INDUSTRY GROWTH。
THE TERMINAL HARDWARE WILL MOVE FROM A SIMPLE "TOOL APPLIANCE" TO A "NATIVE AI DESIGN". A NEW GENERATION OF AI MOBILE PHONES, PCS AND XR DEVICES WILL BE COMBINED IN DEPTH WITH LARGE, MULTI-MODULAR MODELS TO PRODUCE NEW, NON-SYMBIOTIC CONSUMPTION SCENARIOS THAT WILL BRING QUALITATIVE LEAPS IN INDIVIDUALIZED EDUCATION, HEALTH MANAGEMENT AND RECREATIONAL EXPERIENCES。
AI SMART
"PHYSIC AI" IS INTEGRATED WITH "PHYSICAL INTELLIGENCE" AND PROMOTES INTERACTION BETWEEN ROBOTS AND REALITY DEPTHS。
IN 2025, MORE TECHNOLOGICAL BREAKTHROUGHS WERE ACHIEVED BY SMART ROBOTS. THROW AWAY REMOTES, AND CARRY THE WORLD ' S FIRST PERSONAL-SHAPED ROBOT GENERAL VISUAL PERCEPTION SYSTEM, WITH SMART ROBOTS ABLE TO RUN AUTONOMOUSLY ON SPORTS TRACKS; MOVING FROM PROTOTYPES TO MASS PRODUCTION, SMART ROBOTS WALK INTO REAL SCENES SUCH AS INSPECTIONS, SERVICE OFFICES, FACTORIES, AND OLD-AGE HEALTH CARE, AND TAKE HUNDREDS OF MILLIONS OF DOLLARS IN ORDERS. IN 2025, THE SIZE OF CHINA ' S SMART MARKET IS PROJECTED TO BE $5,295 MILLION, OR APPROXIMATELY 271 TP3T WORLDWIDE; FROM ITS OWN BODY TO ITS BRAIN, SMART ROBOTS ARE ACCELERATING THEIR ITERATIVE EVOLUTION. “PHYSICO AI” IS IN DEEP INTEGRATION WITH “SMART” TO DRIVE INTELLIGENT ROBOTS FROM A STRUCTURED ENVIRONMENT TO A MORE COMPLEX OPEN SCENE. IN 2026, SMART ROBOTS ARE EXPECTED TO LAUNCH LANDMARK PRODUCTS IN AREAS SUCH AS MANUFACTURING, WAREHOUSING, HOME SERVICES AND TO BEGIN THE SCALE-UP PILOT PHASE TO ACHIEVE DEEP INTERACTION WITH THE REAL WORLD。
Further breakdown and deepening of professional areas
"AI+ Science" (AI for Science) produces subversive results in basic disciplines。
SCIENCE INTELLIGENCE DRIVES THE NEXT GENERATION OF AI TO EVOLVE, ENABLING A FULL-SCALE SCIENTIFIC PARADIGM CHANGE. AI LARGE MODELS ARE COMBINED WITH SCIENTIFIC COMPUTATION DEPTHS AND BEGIN TO GENERATE ASSUMPTIONS, DESIGN EXPERIMENTS AND VALIDATE THEM. ACCELERATE THE PROCESS OF “0 TO 1” IN LIFE SCIENCES (E.G. ANTIBODY DESIGN, NEW DRUG MOLECULES), MATERIAL SCIENCES, ASTROPHYSICS, ETC. BY STRENGTHENING THE ALGORITHMS, THE ALGORITHMS AND THE THREE CORE ELEMENTS OF THE DATA IN SYNERGY AND INNOVATION, THE COUNTRY HAS BUILT A NETWORK OF INTELLIGENT SCIENTIFIC TOOLS IN THE FRONTIER AREAS OF PROTEIN STRUCTURE PREDICTION, QUANTUM SIMULATION, MATERIAL GENOMES, AMONG OTHERS, TO PROMOTE A SHIFT IN SCIENTIFIC RESEARCH FROM LINEAR GROWTH TO INDEXING。
Cross-integration of frontier areas
BRAIN INTELLIGENCE AND CROSS-CUTTING DISCIPLINES ACCELERATE INNOVATION. BRAIN SCIENCE EXPLORES THE NATURE OF COGNITION, CONSCIOUSNESS AND INTELLIGENCE, WHICH IS THE “END FRONTIER” IN WHICH HUMANS UNDERSTAND NATURE; BRAIN-TYPE TECHNOLOGIES ARE INSPIRED BY THEM TO DEVELOP INTELLIGENT SCIENCE AND TECHNOLOGY SUCH AS BRAIN ALGORITHMS, DEVICES AND ROBOTS. BRAIN SCIENCE DRIVES FORWARD BRANCH DEVELOPMENT IN BIOIMAGE, DATA SCIENCE, AND BRAIN TECHNOLOGY PROMOTES THE OPTIMIZATION OF ARTIFICIAL INTELLIGENCE ALGORITHMS, ENABLING AUTOMATIC DRIVING, SMART MEDICINE, ETC. THE DEEP INTEGRATION OF BRAIN SCIENCE WITH AI WILL DRIVE A BREAKTHROUGH IN HARDWARE AND ALGORITHMS FOR SUBVERSIVE TECHNOLOGIES SUCH AS PULSE NEURAL NETWORKS, NEUROMORPHOLOGICAL COMPUTATION。
Evolving energy issues
GREEN AI GETS ATTENTION。
AI ' S HUGE ENERGY CONSUMPTION AT THE DATA CENTRE WILL ACCOUNT FOR A SIGNIFICANT PART OF THE INCREASED GLOBAL DEMAND FOR ELECTRICITY, RAISING CONCERNS ABOUT ENERGY SUPPLY AND THE ENVIRONMENT, AND WHERE LOW-COST, RELIABLE AND CLEAN ELECTRICITY CAN BE PROVIDED ON A LARGE SCALE, WILL HAVE A STRUCTURAL ADVANTAGE IN ATTRACTING INVESTMENT IN ARTIFICIAL INTELLIGENCE。
THE FORUM “AI AND GREEN LOW CARBON DEVELOPMENT” OF THE WORLD CONGRESS OF ARTIFICIAL INTELLIGENCE PROPOSED A CHINESE APPROACH TO THE “AI ENERGY PARADIGM”. CHINA'S ELECTRICITY CONSTRUCTION GROUP HAS RELEASED THE RESULTS OF ITS “CARBON-SMART HUB” AIMED AT CREATING A “DIGITAL BASE” AND “CENTRE SYSTEM” FOR A FUTURE GREEN SMART ECONOMY, INTEGRATING THE COORDINATED MANAGEMENT AND GLOBAL OPTIMIZATION OF PREVIOUSLY INDEPENDENT “ENERGY FLOWS, CARBON FLOWS, DATA FLOWS”, AND ALLOWING ENERGY EXPERTS AND BUSINESS EXPERTS TO GIVE INSTRUCTIONS IN THEIR OWN LANGUAGES TO WORK WITH ARTIFICIAL INTELLIGENCE ON COMPLEX GREEN LOW-CARBON OPTIMIZATION TASKS, MARKING A CRITICAL STEP FROM DECENTRALIZED APPLICATIONS TO A SYSTEMATIC TOP-LEVEL DESIGN。
BY DEVELOPING MORE EFFICIENT MODELLING STRUCTURES, USING CLEAN ENERGY COMPUTING CENTRES, AND EXPLORING NEW ENERGY MODELS SUCH AS SMALL MODULAR NUCLEAR REACTORS (SMRS), INDUSTRY WILL BE COMMITTED TO ACHIEVING A BALANCE BETWEEN COMPUTING GROWTH AND CARBON EMISSION CONTROL。
Security and confrontation heat
SECURITY AND GOVERNANCE WILL BE IMPORTANT GUARANTEES FOR AI DEVELOPMENT。
As data poisoning, counter-attacks and deep forgery become a real threat, safety protection will become an endogenous requirement for AI model development. The official launch of the 2.0 version of the Artificial Smart Security Governance Framework at the National Cyber Security Awareness Week main forum in 2025 marked the new phase of AI governance from principle-building to systematization, dynamicization and standardization. Faced with the risks of data misuse, algorithm discrimination and uncontrolled modeling as a result of the rapid development of the generation AI, the new version of the framework enhances risk classification, adds a new derived security dimension, promotes the pre-emptive and ethical aspects of the process, and achieves synergy between technical, ethical and social governance. Through hierarchical regulation and institutional interfaces, enterprises are provided with a path of compliance that will help them to develop their industries in a healthy and orderly manner, while at the same time demonstrating China's “development with security” governance wisdom in AI。
CODES OF GOVERNANCE AND TECHNICAL TOOLS FOR AI ETHICS, PRIVACY AND SAFETY WILL BE ACCELERATED AND IMPROVED IN TANDEM, AND WHILE ENCOURAGING TECHNOLOGICAL INNOVATION, SECURE LINES WILL BE ESTABLISHED TO ENSURE THE HEALTHY AND ORDERLY DEVELOPMENT OF THE ARTIFICIAL INTELLIGENCE INDUSTRY。