As the upsurge of generative AI gradually fades, global artificial intelligence competition has entered a new stage. This competition is no longer a contest of a single technology, but has evolved into a comprehensive systematic game covering technological R&D, industrial ecology, talent reserves, rule-making, and even national sovereignty. At the beginning of 2026, the pattern where China and the United States lead as the “two superpowers” and many countries seek breakthroughs through “sovereign AI” has become increasingly clear. The integration of hardware and software, ecological stratification, and diversified governance have become the core characteristics of the new pattern, profoundly reshaping the global power structure of science, technology, and economy.

The solidification of the pattern and the emergence of variables constitute the basic landscape of global AI competition. Currently, the “two superpowers and multiple strong powers” competition pattern has been fully solidified, with China and the United States occupying the industrial high ground by virtue of their respective advantages. The United States takes “AI for Science” as its core strategy, mobilizes federal scientific resources through the “Genesis Program”, and focuses on technological breakthroughs in basic research and national security fields. The full-stack ecology built by its technology giants from chips and cloud platforms to foundation models has made it firmly occupy the position of the world’s AI “infrastructure operator”. In 2024, U.S. private investment in AI reached as high as 109.1 billion U.S. dollars, accounting for a dominant position globally. China, on the other hand, promotes the in-depth integration of technology and the real economy through the “Artificial Intelligence +” initiative, forming a complete industrial chain covering the basic layer, technology layer, and application layer. As of September 2025, the number of AI enterprises exceeded 5,300, accounting for 15% of the global total. At the same time, it actively exports governance solutions and has become an important participant in global AI governance.

While China and the United States take the lead, the wave of “sovereign AI” has become an important variable to break the pattern. Excessive reliance on AI services from U.S. technology giants has exposed various countries to triple anxieties: economic costs, cultural biases, and data security. Non-English-speaking countries are even burdened with additional “token taxes” and cultural adaptation challenges. Against this background, many countries from Singapore and India to Saudi Arabia have elevated “sovereign AI” to a national strategic height. Saudi Arabia even attempts to transform itself into a global “computing power refinery” by virtue of its energy advantages. Medium-sized powers such as the United Kingdom and Canada have chosen to join forces, exploring the establishment of transnational consortia to break the monopoly barriers of Chinese and U.S. giants through collective strength, forming a diverse force in global AI competition.

The upgrading of competition dimensions is the most distinctive feature of the new pattern. AI competition has evolved from a simple contest of algorithms to a “hardware-software composite” systematic war covering chips, algorithms, talents, and supply chains. At the hardware level, the mass production of TSMC’s 2nm process has set a new benchmark for computing power, and its capacity allocation directly affects the global AI R&D process. Enterprises with advanced process adaptation capabilities will establish insurmountable barriers. At the software level, the dispute between open-source and closed-source paths continues to intensify. U.S. giants adhere to the closed-source model to build technical barriers, while China forms a “crowdsourcing effect” for technological iteration through open-source platforms such as Alibaba’s ModelScope. As of July 2025, the number of open-source models on the platform exceeded 70,000, with users exceeding 16 million.

Talent competition has become the core grasp of systematic competition. OpenAI offers a sky-high monthly salary of nearly 20,000 U.S. dollars to top interns, and Google focuses on recruiting student researchers in fields such as silicon engineering and AI hardware, indicating that talent competition has extended from mature scientists to top students, with a “hardware-oriented” focus trend. Cross-disciplinary talents who can understand algorithm needs and design dedicated chips have become scarce resources. Various countries have relaxed visa policies to attract global intelligence, and the thickness of talent reserves directly determines the competitive potential in the next decade. At the same time, the global industrial chain is forming a hierarchical structure based on the AI ecosystem: “pioneer countries” lead the layout, “potential countries” embed and connect, and “backward countries” face the risk of marginalization. ASEAN, India, and other regions have deeply integrated into the global AI division of labor network through differentiated positioning.

Opportunities and challenges coexist, and the development of global AI under the new pattern is full of games and uncertainties. At the industrial level, investment scale continues to explode. In 2024, global AI venture capital exceeded 100 billion U.S. dollars, with 498 “unicorn” enterprises and a total valuation of about 2.7 trillion U.S. dollars. However, the excessive concentration of capital in a few giants not only exacerbates the volatility risk of the capital market but also squeezes the living space of start-ups. At the social level, AI is reshaping the employment structure. The youth employment rate in fields such as software development and customer service has dropped by double digits, and the job market is transforming from a “pyramid type” to a “diamond type”, with the risk of talent gap looming. At the same time, the AI industry has become a new “carbon emission monster”. In 2025, its emissions were equivalent to the annual level of New York City, and its fresh water consumption exceeded the global total demand for bottled water, posing severe challenges to sustainable development.

The game at the governance level is even more intense. The United States pursues technological blockades with a zero-sum mindset, hindering technological circulation through measures such as chip export controls, which exacerbates the risk of global technological fragmentation. Europe has built a strict regulatory framework relying on the General Data Protection Regulation (GDPR). China has released the “Global Initiative on Artificial Intelligence Governance” to promote the formation of an inclusive and prudent governance consensus. Based on differences in supply chain constraints and data regulations, different regions are gradually forming differentiated technical paths and application ecosystems, and a geoeconomic and technological system is taking shape.

Looking to the future, the core of global AI competition will focus on systematic integration capabilities and ecological inclusiveness. Countries and enterprises that can achieve the coordinated optimization of hardware, software, algorithms, and talents will lead the development direction of the next generation of AI technology. The balance between the expansion of open-source ecosystems and “sovereign AI” will determine the degree of inclusiveness of global AI development. This competition is no longer a zero-sum game of “winner-takes-all”, but requires finding a balance between technological breakthroughs and risk control, hegemonic competition and win-win cooperation, and efficiency improvement and social equity.

The new pattern of global AI competition is essentially a concentrated manifestation of the transformation of productive forces and the adjustment of production relations. Whoever can solve core problems such as talent shortages, energy consumption, and governance imbalances, and whoever can build an open and inclusive technological ecosystem and governance system, will take the initiative in this competition that defines the future. For all countries, only by relying on their own advantages and embracing collaborative innovation can they seize opportunities in the AI wave, achieve high-quality development, and jointly promote artificial intelligence to become a powerful driving force for human progress.