The escalating artificial intelligence (AI) rivalry between the United States and China reached a critical juncture in July 2026, with the Asia-Pacific Economic Cooperation (APEC) Digital Weeks in Chengdu, China, serving as a prominent battleground. While American officials and tech giants projected a subdued presence, China seized the opportunity to champion its open-source AI models and extend its technological influence across Asia, underscoring a strategic divergence that will likely redefine the global AI landscape. The event, held in the southwestern Chinese metropolis, highlighted the contrasting strategies of the two global powers vying for dominance in the world’s largest and most dynamic continent.
The Stakes: A Race for AI Hegemony in Asia
The foundational premise of this intensifying competition, as articulated by industry experts, is a strategic imperative for the United States to prevent China from becoming the preeminent AI supplier for Asia and, by extension, the world. Gary Dvorchak, managing director at The Blueshirt Group, posited, "The American strategy is to stop China from becoming the leading AI supplier for the rest of Asia … and frankly the whole world." Dvorchak acknowledged China’s competitive advantage in offering cheaper alternatives, yet emphasized the U.S.’s current lead in providing a more comprehensive AI solution, spanning from advanced semiconductor chips to sophisticated AI models. This "full-stack" advantage has historically been a cornerstone of American tech leadership.
However, the U.S. faces significant challenges in translating this technological superiority into market penetration, particularly in a region eager for cost-effective and adaptable solutions. The APEC Digital Weeks in Chengdu starkly illustrated these difficulties, revealing a strategic gap between Washington’s ambitions and its on-the-ground execution.
A Subdued American Presence in Chengdu
Despite earlier pronouncements by U.S. officials and business representatives to APEC highlighting AI’s importance in promoting the Chengdu event, the American footprint at the Digital Weeks was notably modest. Public records and social media updates from U.S. TradeGov offered few details of active involvement, contrasting sharply with past engagements. This understated presence followed a period of policy volatility, including Anthropic’s flip-flop on its Fable AI model release due to abrupt U.S. policy changes. Concurrently, new Chinese AI models, such as Moonshot AI’s Kimi K3, had recently launched with similar capabilities at significantly lower price points, further complicating the U.S. sales pitch.
This muted showing stood in stark contrast to the first APEC AI meeting in South Korea the previous summer. During that event, Michael Kratsios, then President Donald Trump’s chief science and technology policy advisor, robustly promoted the U.S. AI Action Plan and the establishment of the American AI Exports Program, as detailed in a White House transcript. The program was designed to facilitate the export of U.S. AI technology and expertise to allied nations, reinforcing American leadership.
Challenges for the American AI Exports Program
The efficacy of the American AI Exports Program itself has come under scrutiny. Earlier in July, Politico, citing three former officials, reported that the Commerce Department had received a less-than-expected 78 applications for the program. This figure suggested a lukewarm uptake, potentially indicating a disconnect between government initiatives and industry interest or perceived market viability.
Despite these reported challenges, Bill Guidera, deputy under secretary for innovation and engagement at the U.S. Department of Commerce, continued to advocate for the program during his address at an APEC High-level Forum on AI in Chengdu on July 24. Organized by China’s cybersecurity regulator, the forum provided a platform for Guidera to call for broader Asia-Pacific partnerships. He emphasized the program’s flexibility, allowing buyers to acquire either a full U.S. tech stack or specific components. Guidera asserted, "It is the brilliant design that shows the strength, security and capability of U.S. AI," portraying the program as a testament to American technological prowess and reliability. The Commerce Department’s International Trade Administration later confirmed Guidera’s participation via a social media post on July 29, and an ITA spokesperson disputed the Politico report, stating the volume of applications "exceeded our expectations," though specifics were not provided. The White House, when contacted by CNBC, did not offer comment.
Beyond official government representation, the presence of American tech giants was also limited. Google and Meta were the only U.S. companies observed by CNBC with exhibition booths as of July 23, standing alongside displays from Thailand and a majority of Chengdu-based Chinese firms. Significantly, both Google and Meta opted to showcase specialized AI applications rather than their flagship large language models (LLMs). Google focused on its molecular AI system, AlphaFold, while Meta highlighted AI tools for small businesses. On July 24, Google’s government affairs vice president, Wilson L. White, made only passing references to Gemini in his speech. This strategic choice underscored a cautious approach, possibly to navigate the complex regulatory and geopolitical landscape in China, or to emphasize niche applications where U.S. technology retains a clear edge.
China’s Assertive AI Diplomacy and Open-Source Strategy
In stark contrast to the U.S.’s reserved approach, China leveraged its role as APEC host to aggressively advance its AI agenda. This year’s APEC meetings coincided with heightened U.S.-China tensions and an intensifying tech rivalry, providing Beijing with a prime opportunity to showcase its capabilities and diplomatic prowess.
A cornerstone of China’s strategy has been its strong advocacy for open-source AI models, a distinct approach compared to the largely closed models favored by many U.S. developers. Chinese President Xi Jinping set the tone earlier, on July 17, at the World AI Conference in Shanghai. There, he announced China’s commitment to provide 5,000 AI training and seminar opportunities to developing countries. He also unveiled plans to establish AI application cooperation centers with Southeast Asian nations and other regions, signaling a clear intent to foster technological ties and export its AI ecosystem. This initiative reflects a broader "AI diplomacy" strategy aimed at solidifying China’s position as a collaborative and accessible AI partner, particularly among emerging economies in the Global South.
Beijing further reinforced this strategy by sending a high-ranking official, Vice Premier Zhang Guoqing, to advocate for developing tech standards with other Asia-Pacific nations at the minister-level APEC Digital Weeks on July 23. This high-level representation underscored the strategic importance China places on shaping global AI norms. Later that same day, the 21 APEC member economies, including the U.S., reached a consensus to back open-source AI with "strong security."
This agreement was a significant diplomatic victory for China. Wei Sun, principal analyst for artificial intelligence at Counterpoint Research, noted, "The ‘endorsement’ of open-source models with strong security assurance gives China’s open-weight strategy greater regional legitimacy, especially across emerging Asian economies where deployment cost and technological sovereignty are major considerations." For many developing nations, open-source solutions offer lower entry barriers, greater customization potential, and a perceived path to greater technological independence, aligning well with China’s collaborative narrative.
The Reality of "Forced Integration" and Localized AI
Despite the fierce competition, experts suggest that the future of AI in Asia may not be a simple division into U.S. or Chinese spheres of influence. Instead, a "forced integration" of technologies from both nations is likely to emerge. Wei Sun anticipates a combination of technologies, driven by practical necessities in a diverse region.
The sheer linguistic diversity of Asia presents a compelling case for this integrated approach. With over 1,300 living languages in Southeast Asia alone, deploying a single, universal U.S. or Chinese AI model is impractical and inefficient. Governments and businesses across the continent are investing billions in developing AI systems specifically tailored to local languages and cultural nuances.
Pak-Sun Ting, CEO of privately funded startup Votee AI, exemplifies this trend. His company, already generating over $10 million in annual revenue, works with at least five governments, including two in Southeast Asia, on localized AI solutions. Ting highlighted that entities in Southeast Asia often combine technologies, utilizing Nvidia chips for AI training due to their industry dominance, while potentially employing other chips for running inference models. He cited Votee’s open-source model for Cantonese speakers, which was partly developed using Alibaba’s open-source Qwen model, illustrating the practical reality of blending technologies from different origins to achieve optimal local outcomes. This pragmatic approach prioritizes functionality and cost-effectiveness over strict adherence to a single national tech stack.
Beyond Chengdu: A Broader Strategic Context
Ultimately, the ability of AI to generate tangible economic returns remains the critical metric, irrespective of its country of origin. Yue Su, principal economist at the Economist Intelligence Unit, observed, "While the U.S. and China are fiercely competing in AI technology and diplomacy through distinct approaches, they ultimately cannot fully decouple from one another." This statement underscores the interconnectedness of the global tech ecosystem, where supply chains and innovation often transcend national borders.
Su also offered context for the U.S.’s relatively light presence at APEC Chengdu, pointing to other significant events that may have diverted American attention and resources. Notably, a San Francisco AI Summit took place concurrently on July 24, signaling a deliberate choice by the U.S. to host its own high-profile discussions. Organized by South Korea’s tech ministry, this summit saw President Lee Jae-myung actively engaging with leading U.S. frontier AI model developers, including Sam Altman of OpenAI, Dario Amodei of Anthropic, and Jensen Huang of Nvidia. Lee’s objective was to build upon South Korea’s ambitious chip and AI megaprojects, valued at trillions of won, by fostering deeper collaboration with American innovators. This simultaneous event in San Francisco suggests a strategic bifurcation of U.S. diplomatic efforts, focusing on established alliances and cutting-edge research at home, while navigating the complexities of engagement in China.
The APEC Digital Weeks in Chengdu, therefore, represented more than just a series of meetings; it was a microcosm of the evolving U.S.-China tech rivalry, demonstrating contrasting diplomatic approaches, technological philosophies, and market realities. As Asia continues its rapid digital transformation, the interplay between these two global powers will undeniably shape the continent’s AI future, likely leading to a multifaceted and integrated technological landscape rather than a clear-cut division.
Jenny Lee contributed to this report.
