China

U.S. lawmakers are investigating DoorDash’s use of Chinese AI models, specifically Kimi K2.6 from Moonshot AI. This inquiry stems from concerns about national security risks associated with American companies relying on AI developed in China. DoorDash stated its commitment to American AI innovation and looks forward to engaging with committees on responsible AI use. The investigation is part of a broader congressional effort to assess the risks of adopting Chinese AI technologies.

U.S. tech giants urge Washington to adopt a robust, open-source AI strategy to counter China’s growing influence. While the U.S. focuses on domestic chip manufacturing, it lacks a clear approach to open-weight AI models. Open models offer cost reduction, data control, and transparency benefits. China is actively advancing its open-source AI, potentially dominating global infrastructure. The U.S. must invest in open AI development and security to maintain competitiveness, rather than imposing bans, to avoid hindering innovation and global adoption.

The FCC has restricted imports of advanced foreign-produced robots, including humanoids, citing cybersecurity concerns. China’s Commerce Ministry strongly criticized the move, accusing the US of disregarding Beijing’s stance and threatening retaliation. This action impacts China’s leading humanoid robotics sector, with companies like UBTech and Unitree facing market challenges. China may retaliate by restricting rare earth sales or market access for US firms. The situation escalates US-China trade and technology tensions amid an AI race.

In the AI race, the US aims to limit China’s global AI supply dominance. At APEC’s Digital Weeks, US participation was understated, with companies like Google and Meta showcasing specialized, not flagship, AI. China, conversely, is pushing open-source models and building regional AI ecosystems. While competition is fierce, complete decoupling is unlikely, with Asia favoring customized solutions and potential technological convergence.

China reports a breakthrough in mass-producing immersion DUV chipmaking tools, prompting investor concern. While this poses a potential challenge to ASML’s dominance, experts suggest its impact may be limited to lower-end chips. Key hurdles for China include achieving yield parity and scaling production, areas where ASML holds a significant advantage due to decades of refinement and deployment, particularly in advanced EUV lithography.

China’s CXMT DRAM maker’s IPO is generating massive speculation in a nascent derivatives market, with offshore investors driving its perceived valuation to over $425 billion, far exceeding its offer price. This surge is due to limited access to the Shanghai STAR Market listing for international and mainland retail investors. Crypto traders on Hyperliquid are pricing the company at stratospheric heights, creating an arbitrage opportunity and highlighting the evolving role of decentralized finance in price discovery for inaccessible assets.

SenseTime launches the “Galaxy Project” with nearly 20 partners to expand China’s AI chip infrastructure. The initiative focuses on a closed-loop system integrating chip development, ecosystem partnerships, and AI computing deployment. Key strategies include improving token processing efficiency, enabling multi-chip adaptability, and pioneering energy efficiency metrics. The project also explores advanced computing like space and quantum. While ambitious, claims of performance and cost-effectiveness require independent verification.

U.S. Treasury Secretary Scott Bessent is scrutinizing Chinese AI models for potential intellectual property theft, citing “watermarks” from American LLMs. This concern arises as Chinese open-weight models gain traction, challenging U.S. AI dominance. Bessent warned of sanctions for IP violations, particularly through “distillation.” The US plans AI dialogues with China in September, highlighting the global race and the need for fair competition and IP protection in the rapidly evolving AI sector.

ASML, a key supplier of chipmaking equipment, faces geopolitical pressure amid booming AI demand. While China remains a crucial market, the US is pushing for stricter export controls on less advanced DUV lithography machines. ASML, which doesn’t export its most advanced EUV technology to China, has raised its guidance due to strong customer demand for AI chips. Potential US legislation like the MATCH Act could impact its orderbook, forcing ASML to balance global demand with international relations.

At the World AI Conference in Shanghai, President Xi Jinping invited the Global South to collaborate on AI development, pledging 5,000 training opportunities and enhanced cooperation with ASEAN, the Arab States, and the African Union. He advocated for a “symphony of international cooperation” and stressed AI must be secure and human-controlled. This comes as 29 countries signed an agreement to establish the World Artificial Intelligence Cooperation Organization (WAICO) in Shanghai.