武林外传
China advances global AI governance on all fronts, turning principles into deeper, more concrete action_我的网站
一 |
When machines begin to think, how should humans coexist with them? When algorithms participate in decision-making, how can security be ensured? When technology challenges ethics, how can governance keep pace? As the divide continues to widen, how can inclusiveness be achieved?
At the opening ceremony of the 2026 World AI Conference (WAIC) and High-Level Meeting on Global AI Governance, Chinese President Xi Jinping pointed out that we should always guide AI development with human wisdom and international consensus, so that AI can truly become a mighty force that increases the well-being of humanity and advances human civilization.
In the intelligent era, we share the same planet and must pull together. Standing at the forefront of the times and shouldering the responsibilities of a major country, China is advancing global AI governance in an all-round way, making it deeper and more substantive. From laying down the opening strokes through its own development, to the "broad-brush freehand" of top-level design, and then to the "fine-brush painting" of detailed implementation, China is working to make the dividends of technology benefit the world and to make the bottom line of security clearly visible. This is China's commitment to the world, and also humanity's answer to the future.
Building a solid foundation for governance through strength at home
Not long ago, Kimi K3, a new-generation large AI model developed in China, was released as open source and was made available for anyone to download. With 2.8 trillion parameters, it is currently the world's largest open-source model in terms of parameter count.
Kimi K3 is one member of the "China team" in the AI field that have attracted global attention. Six of the world's top 10 applicants for generative AI patents are headquartered in China. The Financial Times has noted that Chinese companies are joining the global front ranks in AI innovation while outperforming their peers in cost efficiency.
Since the 18th National Congress of the Communist Party of China (CPC), the Party Central Committee with Comrade Xi Jinping at its core has attached great importance to AI development, improving top-level design, strengthening work arrangements, and promoting an overall and systematic leap in China's comprehensive AI strength.
Data show that the adoption rate of AI technology among China's manufacturing enterprises above designated size has exceeded 30 percent, the core industries of the intelligent economy have grown to more than 1 trillion yuan (about $139 billion), the number of generative AI users has surpassed 600 million, and China has become the world's largest holder of AI patents. China has achieved notable results in AI technological innovation and industrial applications, with its overall strength significantly enhanced.
Standing at the new starting point of the 15th Five-Year Plan period (2026-30), China is consolidating the foundation for innovation through greater self-reliance and strength in science and technology and embracing the global opportunities through the "AI Plus" initiative. China is entering a prime opportunity to develop its AI.
This confidence comes from continuous breakthroughs in core technologies in key fields.
Zhangjiang Science City, where the Zhangjiang Science Hall exhibition area of the 2026 World AI Conference is located, is home to hundreds of companies for chip designing, manufacturing, packaging and testing, with the scale of the industry accounting for about one-fifth of the national total. In the chip sector, which has been affected by external shocks, China has not only blazed a new path through independent innovation and self-reliance, but has also firmly kept the initiative in AI development in its own hands.
"Computing power is the cornerstone of AI development and governance. It concerns not only technological breakthroughs, but also strategic initiative. Only by basing development on independence and controllability, while ensuring inclusiveness and accessibility, can we continuously strengthen the foundation for AI development," Moore Threads founder Zhang Jianzhong said.
This confidence comes from an increasingly vibrant innovation ecosystem.
With strong support from scientific and technological innovation, AI is being integrated across all areas of China's economy and society, entering ordinary households and bringing tangible benefits to people's livelihoods, making "Intelligent Manufacturing in China" another shining calling card of Chinese modernization.
In Shunde, South China's Guangdong Province, and Wuhan, Central China's Hubei Province, after Midea's "AI agent factories" introduced SmartMore's AI visual reasoning platform, the pace of inspection came fully in line with the production rhythm. AI now directly participates in production decision-making, rapidly turning large models into productive capacity.
In Nanchong, Southwest China's Sichuan Province, and Bozhou, East China's Anhui Province, iFLYTEK Healthcare's "AI Medical Assistant" has significantly improved the efficiency and standardization of medical record writing through functions such as structured consultations and intelligent prompts.
A number of terminal device companies are relying on a well-developed industrial chain support system to accelerate the launch of new-generation products such as smartphones, intelligent robots and smart glasses, enriching market choices and creating new hot spots in information consumption.
On August 12, Chinese smartphone brand Honor launched the world's first Robot Phone.
"China has rich and diverse production and everyday-life scenarios as well as integrated software-hardware engineering capabilities, so this has provided excellent conditions for AI to be efficiently translated from needs into actual experience," Honor CEO Li Jian said.
This confidence comes from institutional guarantee that coordinates development and security.
Recently, Dongbi Tech Data, together with the School of Digital Economy at Shanghai University of Finance and Economics, released the Global Large Language Model Safety Evaluation Report (2026). Using 313 high-risk science and technology questions as the test set, the report assessed the security prevention capabilities of 38 large language models from China and abroad while helping users obtain, understand and use scientific and technological knowledge.
"Currently, society as a whole is paying close attention to issues such as content security and data security arising from the large-scale applications of AI. Balancing development and security, and advancing scientific and technological innovation in parallel with the rule of law in science and technology and the building of ethics and integrity, will further guide large models and the entire AI industry toward positive and beneficial development and sustainable growth," said Zhu Xufeng, dean of the School of Public Policy and Management of Tsinghua University.
In recent years, the State Council's Guidelines on Further Implementing the "AI Plus" Initiative have explicitly called for establishing and improving systems for AI technology monitoring, risk early warning and emergency response. The Implementation Guidelines on the Standardized Application and Innovative Development of AI Agents have also made clear that the development of AI agents should adhere to the basic principles of safety and controllability, orderliness and standardization, innovation-driven growth and application-oriented traction. China is accelerating efforts to build laws and regulations, policies and institutions, application standards and ethical guidelines that are suited to the evolving trends of AI, strengthening the security foundation for Chinese modernization and contributing Chinese wisdom and Chinese approaches to global science and technology governance.
Guiding AI governance with a global vision
A recent commentary from The New York Times likened the development of cutting-edge AI models to two divergent paths.
"The United States is behaving as though frontier A.I. models could be the equivalent of nuclear weapons, offering an immensely destructive form of power that requires a careful and jealous watch… Meanwhile, China treats the models that its companies produce as the equivalent of nuclear energy, a lower-risk technology to be shared and commercialized to win friends," read the NYT article.
At the heart of these two approaches lies a fundamental difference in how AI is perceived: Is it a strategic asset to be monopolized in a race for dominance, or a public good that should serve humanity?
China's answer is clear and unwavering: Put people first, ensure AI is used for good, and make it a powerful force for shared prosperity and common security.
Xi called for joint efforts to build a just and equitable system for global AI governance, the Xinhua News Agency reported.
This reflects a deeper understanding of technological and historical trends, promoting shared benefits through consultation and wisdom through good governance. It also demonstrates China's commitment to global AI governance, ultimately serving the vision of a community with a shared future for humanity.
Embracing openness and win-win cooperation to drive innovation and empower industries.
China has emerged as one of the world's most active and fastest-growing hubs for open-source AI models. Chinese open-source AI models have been downloaded more than 10 billion times worldwide, while open-source communities such as AtomGit have brought together more than 11 million registered users.
From topping global large-model leaderboards in certain performance metrics, to steadily advancing multimodal capabilities, and to evolving from single intelligent agents into agent clusters, China's large-model innovation and iteration have been accelerating at an unprecedented pace. This has not only given rise to a diverse array of intelligent application scenarios and unlocked new potential for domestic economic growth, but has also driven technological inclusiveness through advantages such as high performance and low cost. Chinese large models have become "productivity tools" favored by developers and users worldwide, as well as benchmarks in global model evaluations and important reference cases for overseas research institutions.
Openness and open source are more than just a technological development model. They embody a broader philosophy of collaboration and coexistence, as well as a commitment to sharing the fruits of development.
China has launched the AI+ International Cooperation Initiative, calling upon all countries to proactively carry out AI+ campaigns in light of their national conditions, enhance policy exchanges and practical cooperation, and share best practices and solutions on the basis of mutual respect and mutual benefit.
The Action Plan for AI Cooperation and Development sets out eight areas of action, including sharing open-source ecosystems, jointly developing rules and standards, strengthening cooperation on safety governance, and ensuring AI is used for good.
China has launched "Fenghe," a large language model for AI-powered meteorological services, and initiated a global open-source program.
Strengthening risk awareness to ensure AI remains safe, controllable and trustworthy.
With a deep understanding of the trends and logic of AI development, we are continuously improving laws, regulations, policies, mechanisms, application norms, as well as ethical principles to make sure that AI is safe, secure and controllable, and that this fine steed of AI gallops with both speed and stability, Xi said, according to Xinhua.
China has put forward the Global AI Governance Initiative, advocating risk-based assessment, stronger laws and regulations, and the development of ethical guidelines. It has called on countries to jointly prevent risks and continuously improve the safety, reliability, controllability and fairness of AI technologies.
China has also released versions 1.0 and 2.0 of its AI Safety Governance Framework, embedding safety requirements throughout the entire process of model development, deployment and application.
The Global AI Governance Action Plan has identified "safety and controllability" as one of its core principles; An international action plan on AI ethics governance has contributed a framework that is both principled and actionable to global AI governance... From defining directions to charting pathways, China continues to provide international public goods, ensuring that the "swift horse" of AI runs both fast and steadily.
Fostering inclusiveness and mutual learning to build a diverse and harmonious global civilization.
Dujiangyan tamed the Minjiang River without dams, nourishing the Chengdu Plain for over 2,000 years. Confucius taught that "Gentlemen seek harmony but not uniformity." Civilizations have their own paths, best pursued through mutual respect and coexistence.
Drawing on such Eastern wisdom, Lu Zilong, deputy chair of the World Intellectual Property Organization supervision committee, interprets China's coherent approach to AI governance.
Today, the Eastern wisdom is being translated into concrete initiatives. The "cultural interaction engine," developed by the institute for Long-term AI Research Center and the Institute of Automation of Chinese Academy of Sciences, uses AI to uncover links between world cultural heritage sites, presenting the shared values and diversity of human civilization to users in more than 170 countries and regions.
The Milan Winter Olympics has launched the first official LLM in Olympic history, powered by Qwen, a Chinese large model, serving both professional operations and public services, a new era of smart winter sports experiences.
Chinese and Malaysian companies have jointly established a multilingual AI translation center for ASEAN, supporting translations across more than 130 languages and enabling film and television content to be translated in as little as 30 minutes.
From cultural dialogue to sports, from language to human connections, Chinese AI is bridging distances, allowing civilizations to flourish through inclusivity and be revitalized through exchanges.
Promoting solidarity and improving global governance to prevent new historical injustices in the AI era.
During the 2026 WAIC and High-Level Meeting on Global AI Governance, the World Artificial Intelligence Cooperation Organization (WAICO) was established, with 38 countries becoming its founding members.
The blueprint laid out a year ago has now borne fruit. As the world's first intergovernmental international organization on artificial intelligence, WAICO is committed to AI for good and inclusive benefits, adheres to the purposes and principles of the UN Charter, and is open to all countries, charting a new course and coordinates for global governance in the age of intelligence.
China has sought to translate its sense of responsibility into concrete action. By upholding multilateralism, bringing the international community together and continuously seeking the broadest possible common ground on global AI governance, China is working to ensure that AI can truly become a valuable common asset of all humanity.
Looking to the future, fulfilling the responsibility of governance through action
Looking back at the summer of 1956, the spark of "artificial intelligence" was quietly kindled at a workshop at Dartmouth College in the US.
Over the past 70 years, AI scientists and researchers around the world have explored the unknown, moved forward through twists and turns, and achieved breakthroughs through perseverance, continuously expanding the boundaries of technology.
History has witnessed Chinese science and technology carving furrows out of barren ground and setting off from a faint glimmer of light. China has progressed from following behind to running alongside others and then to gaining the initiative on more tracks. Every rut and every milestone along the way has deepened its understanding of the power of technology, while also prompting clearer reflection on where the road ahead will lead and how far it can go.
Inclusive development for the benefit of all is the right path forward, and it is a path that grows broader as we travel further.
Remote control of pond aeration, round-the-clock online monitoring of water-quality parameters, precise feeding by drones... In Takeo Province in southern Cambodia, with support from an integrated smart farm development program implemented by China, an intelligent rice-fish farming model has significantly improved the survival rate and quality of local giant freshwater prawn seedlings. Annual income per hectare has increased from about $1,000 to more than $3,000, bringing real increases to the incomes of local farmers.
In Saraburi Province, Thailand, the Chinese-developed "YUKON" system has equipped a traditional mining area with a "digital brain." From excavators to mining trucks, and from power monitoring to transport routes, this parallel mining operating system, based on an AI large model, can make the optimal allocation of equipment across the entire mine within milliseconds. Efficiency has risen by 16 percent and electricity consumption has fallen by 20 percent.
SCG International General Manager Jirapat Janjerdsak said, "China's AI technology has shown us the possibility of achieving a mine with 'zero casualty, high efficiency and low emissions.'"
Along the Red Sea coast, as the tide recedes, sediment samples are collected into containers. Through China's AI-powered marine drug intelligent screening platform, they become a "seed bank" for new antibacterial agents.
At Djibouti's Doraleh Multi-Purpose Port, extreme weather such as high temperatures, short bursts of heavy rain and sudden strong winds frequently causes disruptions. MAZU, an AI-powered meteorological early-warning solution from China, has helped establish a full-chain early-warning system covering "monitoring, forecasting, warning, services."
Helping safeguard countless homes and protect peace and safety far and wide, Mohamed Ismail Nour, director general of the Djibouti National Meteorological Agency, said, "Djibouti hopes to steadily advance its economic and social development by drawing on Chinese technologies."
To bridge the digital and intelligent divide, it is better to teach people how to fish than simply give them fish.
Data from the International Telecommunication Union show that about one-third of the world's population still has no access to the internet. The World Bank's Digital Progress and Trends Report 2025: Strengthening AI Foundations shows that low- and middle-income countries lag far behind high-income economies in both the number of AI systems and available computing capacity; a survey by the African Data Centres Association has found that most of the world's data center capacity is concentrated in developed economies, with Africa accounting for less than 1 percent.
Without coordinated intervention, disparities in AI development may further evolve into a digital and intelligent divide, or even a North-South divide, posing new challenges to global development.
China has held AI capacity-building workshops, launched a special cooperation program on AI science and technology innovation under the China-proposed Belt and Road Initiative, developed the "Digital South" brand under the Global Development Project Pool and established the China-ASEAN AI Industry Innovation Center… A series of pragmatic measures have crossed mountains and seas, using Chinese expertise to help countries, particularly those of the Global South, strengthen their capacity building.
The great road takes us far when traveled together; working together, we embark on a new journey. China has announced that over the next five years, it will provide developing countries with 5,000 opportunities in AI training and seminar programs; develop international AI application cooperation centers with the ASEAN, the League of Arab States, the African Union, the Community of Latin American and Caribbean States, the Shanghai Cooperation Organisation and BRICS; and promote the applications of the MAZU AI-powered meteorological early-warning solution in 30 countries.
Standing at a new historical starting point, China is willing to work with all parties with a more open approach, more pragmatic action and a longer-term vision, to seize the opportunities and address the challenges brought by AI development, creating a better future for human society.
This was compiled and translated by the Global Times based on an article originally published on Page 1 of the People's Daily on August 13, 2026.
。 俄乌冲突持续多年以后，现代战场上一个很朴素的问题越来越显眼：坦克、火炮、无人机再先进，也得有人送油、送弹、修车、抬伤员。2026年《解放军报》谈现代战场工程防护时还专门提到，俄乌等近期冲突中，卫星侦察、无人机和精确制导武器的广泛使用，让交通线、后勤节点和固定设施越来越容易暴露。七十多年前的朝鲜战场没有今天这么多无人机，美军的飞机却已经把这个问题摆到了志愿军面前。
二 |
到了1986年10月，洪学智以解放军总后勤部部长身份访问美国。晚宴上，美军太平洋舰队司令莱昂斯突然问他：“将军是哪个军校毕业的？”洪学智的回答很有意思，他说自己的“学校”就是美国空军。中国军网后来解释这句话时写得很直白：当年的美军空袭，硬是把一个原本主要带兵打仗的将领逼成了现代军事后勤专家。
三 |
美国人几十年后还问这一句，并不奇怪。 因为他们当年真没弄明白，那条运输线为什么一直断不了。 洪学智去朝鲜以前并不是专门搞后勤的干部。1950年10月，他作为中国人民志愿军副司令员入朝，分管司令部、特种兵和后勤等工作。 当时志愿军甚至没有后来那种独立完整的战区后勤体系，很多供应还需要依靠国内和东北方向组织。 以前在国内打仗，部队有时能够就地筹粮，武器装备还可以大量依靠战场缴获。
四 | 到了朝鲜，这套办法一下遇到了麻烦。当地经过战争破坏，几十万人的军队不可能全靠附近村庄解决给养；美军又拥有很强的空中力量，缴获的重装备还没来得及利用，就可能遭到飞机轰炸。弹药型号也越来越复杂，靠“打到什么用什么”已经很难支撑大规模现代战争。
前几次战役打下来，一个限制越来越明显：前线许多粮食和弹药需要战士自己背，一轮大规模进攻往往只能维持约7天。 李奇微后来注意到这个规律，把志愿军这种攻势称为“礼拜攻势”。进攻几天以后，弹药少了，人也疲劳，补给又没及时赶到，美军就抓住间隙反击。 战士敢不敢打，在这里已经不是全部问题。 山那边有十万发炮弹，送不到炮位，等于没有。 后方仓库有粮食，汽车开不到前线，战士照样饿肚子。 洪学智后来一辈子搞后勤，很大一部分起点就在这几个月里。 第五次战役前后，洪学智回国汇报朝鲜战场的供应问题。他提出的已经不是“再多给几辆汽车”这样的小修小补，而是要求把后方重新组织起来：防空、通信、铁道、工兵、兵站、运输不能各干各的，需要有一个机构统一指挥。 中央军委随后决定成立中国人民志愿军后方勤务司令部，洪学智兼任司令员。
五 | 后来洪学智的儿子洪虎回忆，父亲起初还不愿接这个差事，觉得自己本来是军事干部，彭德怀最终还是把任务压给了他。
这个机构成立以后，朝鲜后方才逐渐有了一套完整骨架。
六 | 物资从国内进入朝鲜以后，不再只靠一条道路从头送到底；铁路、公路、兵站开始分段连接，哪里断了就换线路，汽车送不到的地方继续倒运，重要仓库也尽量转入半地下或者隐蔽位置。1951年8月又成立中朝联合铁路运输司令部，后来进一步把运输、抢修、高炮防护结合起来。
洪学智处理的问题很少有什么传奇色彩。 每天都是桥。 路。 油。
七 |
粮。 汽车。 伤员。 可战争拖到几个月以后，这些东西比一场漂亮冲锋更容易决定前线还能不能继续打。 1951年夏天，朝鲜北部又碰上几十年少见的大洪水，公路、铁路和仓库大面积受损。
八 | 紧跟着，美军开始强化针对志愿军后方运输系统的空中封锁，试图从铁路桥梁、车站和公路节点下手，把前线同后方割开。
九 |
中国军网的军史资料记载，1951年8月至1952年6月的“绞杀战”期间，美军一度把大部分空中力量集中用于破坏运输线。 铁路枢纽、桥梁和汽车都是重点目标。战争期间，美军飞机出动规模达到百万架次级别，运输线长期处在空袭压力之下。 打到这个程度，靠一条修得特别结实的公路没有用。 今天修好，明天还会炸。 志愿军后勤后来采取的办法更像是在和飞机比恢复速度。桥炸坏了，抢修；大桥暂时修不了，就搭便桥、绕行线；铁路一边不能走，就想办法利用还能通的一边；公路沿线修汽车隐蔽所，车辆夜间行驶，防空哨负责观察敌机，一有情况就报警。铁路、公路、漕运也能根据情况接起来使用。 有个变化很能看出效果。 战争初期司机夜间不开灯，怕遭敌机发现，一夜往往只能跑30至40公里，撞车翻车还不少。防空哨、隐蔽点以及道路条件逐渐改善以后，到1952年，部分夜间运输距离已经提高到200公里以上。中国军网统计，汽车损失率也从第一年的约40%逐年大幅下降。 这不是某个司机忽然变勇敢了。 路、哨兵、通信、抢修、掩体一起改变以后，汽车才敢跑得更快。 这里还有一组数字，比“钢铁运输线”几个字更有意思。
十 |
根据《解放军报》刊载的抗美援朝后勤资料，1951年1月至1952年2月，美军对志愿军后方的轰炸强度大幅增加，而志愿军物资运输量反而增加了两倍以上。 国防部刊载的洪虎回忆还记录，经过后勤体系不断改进，汽车和物资损失都明显下降，运输效率提高。 铁路上的情况同样如此。1951年7月，美军对铁路的轰炸次数已经是当年1月的数倍，铁路运输量却比1月更高；到1952年，美军轰炸强度继续上升，铁路仍没有被彻底压死。 美国远东空军的人员后来曾把一个问题称为“谜”：中国军队的后勤为什么一直没有中断？相关军史资料甚至记载，美军方面战后很想见见负责这套后勤系统的人。 这件事放到今天看，容易产生一个误解，好像洪学智靠几个奇招骗过了美国空军。 实际规模要大得多。 后方有铁道兵、汽车兵、高炮部队、工兵、仓库、兵站、通信人员，还有大量中朝群众参与抢修和运输。反“绞杀战”时期，仅后勤战线就有十几万人，大规模修路时甚至要动员二线作战部队一起干。洪学智的作用，是把这些过去分散的力量逐渐放进一个统一的保障体系。 这才是他难学的地方。 会修一座桥的人很多。
十一 |
能让桥梁、汽车、防空、仓库、铁路和几十万人的供应同时转起来，是另一回事。
十二 |
洪学智自己文化程度并不高。中国军网介绍1986年那段往事时特别提到，他甚至没有接受过完整的小学教育。可这个人后来两次担任解放军总后勤部部长，1988年再次被授予上将军衔。 到了1986年的美国晚宴上，美军太平洋舰队司令问他是哪所军校毕业的，表面看是普通寒暄。 洪学智为什么偏偏回答“美国空军”？ 因为他1950年进入朝鲜时，也没有现成答案。 美国人先用飞机告诉志愿军，现代战争的后方一样是战场；一座桥、一个仓库、一队汽车都可以成为打击目标。洪学智和志愿军后勤部队随后花了三年时间，学会怎么在这种条件下让前线继续吃饭、开炮、补充人员，把伤员送下来。 七十多年以后，现代战争已经有卫星、无人机和远程精确武器，后勤线面对的危险比当年更多。2026年中国军网讨论现代联合作战时仍把补给线路脆弱、物资消耗强度大、保障体系必须保持韧性列为重要问题。技术换了，仗怎么打也变了，那道基础题没有消失。 洪学智当年没有无人运输车，没有卫星导航，甚至长期没有制空权。 他手里有的，是一群司机、铁道兵、高炮兵、工兵和几条天天挨炸的路。 美国空军把路炸断。
十三 |
晚上有人把它接起来。 第二天再炸。 夜里再修。
十四 |
到后来，美军飞机增加了，送到前线的东西反倒比以前多了。 1986年，美国人终于坐在饭桌前见到了那个当年负责这件事的人。 他们问他在哪儿学的。 洪学智的答案，也就只需要那一句了： 美国空军。
十五 |
返回，查看更多。Current article:http://h5d2.shengyaodanrongchui.buzz/ev2u0m/20260826/4267.html
Published on:07:08:47