Technology competition remained at the center of US-China ties throughout 2025, as Washington expanded export controls in pursuit of AI supremacy while Beijing advanced self-reliance and championed global cooperation in technologies.

Semiconductors and artificial intelligence formed the front line of strategic competition between the world's two largest economies. The United States maintained and imposed restrictions on advanced chips and manufacturing equipment, implementing major expansions of its Entity List targeting Chinese companies.

In March, the US Department of Commerce's Bureau of Industry and Security added more than 80 companies, primarily from China, to its Entity List, focusing on companies in advanced integrated circuits, quantum computing and AI sectors.

In September, the BIS introduced an Affiliates Rule, automatically extending export control restrictions to any foreign entity that is 50 percent or more owned by a party already on the Entity List, though this expansion was suspended for one year in November as part of temporary trade negotiations with China.

Alongside export licensing, the US began restricting outbound investment into China in sensitive technologies. The Treasury Department's final rule on outbound investment screening, which took effect on Jan 2, 2025, targets US investment in Chinese or Chinese-owned companies involved in semiconductors and microelectronics, quantum computing and artificial intelligence.

The US also pressed allies to align with its restrictions. With the "Pax Silica" Declaration signed in December, the US is seeking to form a coalition with partners to build a "trusted" supply chain ecosystem, aimed at countering China's dominance in critical minerals and its growing role in AI and other technology sectors.

Despite tightening constraints on access to leading-edge chips, Chinese firms demonstrated remarkable adaptability and innovation capabilities. Experts questioned the long-term effectiveness of US export controls while highlighting costs to US firms' global competitiveness and broader innovation.

Chinese AI startup DeepSeek exemplified this resilience, shocking the US tech industry early last year with two large language models rivaling dominant US tools at a fraction of the cost and computing power.

Rather than depending on expensive high-end chips, DeepSeek optimized for efficiency, proving that powerful AI could be built via smarter software and hardware optimization.

In December, the firm attracted significant attention with the release of two new AI models that could match or exceed OpenAI's GPT-5 and Google's Gemini 3.0 Pro capabilities.

Similarly, Huawei and Semiconductor Manufacturing International Corporation, both on the Entity List, achieved significant advances in chip production technology.

Huawei's November launch of the Kirin 9030 processor, produced using an evolved version of SMIC's technology, represented "China's most advanced domestic semiconductor manufacturing to date", according to research firm TechInsights.

Analysts have described these developments as both a response to tighter access to foreign chips and a challenge to US-centered digital dominance.

"These developments suggest that export controls may drive innovations in efficiency, with Chinese companies learning to 'squeeze every little bit of IQ out of every flop' under hardware constraints," wrote Mikolaj Barczentewicz, senior scholar at the International Center for Law and Economics, in an analysis.

The Information Technology and Innovation Foundation also said in a May report that "rather than slowing China's ascent, the US government is hastening it", citing Huawei's rapid development of the Ascend chip series as a direct result of US restrictions.

The US and China have also put forward sharply different visions for global AI governance.

In July, the US launched America's AI Action Plan, making it clear it would counter China and compete for global leadership in AI. China released its Global AI Governance Action Plan the same month, emphasizing multilateralism, inclusivity and coordination that includes developing countries.

Room for cooperation

Despite stark contrasts with the two plans, experts identified room for cooperation.

Mark Scott, senior resident fellow at the Atlantic Council's Digital Forensic Research Lab, noted that both nations share "overlapping interests" in AI development where "both sides could benefit from some form of detente".

Both countries are committed to making AI accessible across non-tech industries, potentially opening paths for collaboration on apolitical standards, similar to their 2023 voluntary Bletchley Declaration on AI safety, Scott said. Potential cooperation areas include building mutually accessible datasets for public health and biotechnology, and sharing experiences in developing energy infrastructure to meet the demand of new data centers.

Technology also became interwoven with US-China trade negotiations, increasingly treated alongside tariffs, rare biotechnology and agriculture as bargaining topics. US officials signaled a possible easing of some chip export measures as part of a broader framework tied to implementing a trade truce with China.

In December, US President Donald Trump announced he would allow conditional exports of Nvidia's H200 AI chips to "approved" Chinese buyers in exchange for a 25 percent fee paid to the US government, reversing the near-total block on advanced AI chip exports.

Some analysts viewed the move as part of a broader trade and technology bargain. With trade talks continuing into 2026, semiconductors and AI are expected to remain key items on the agenda and crucial factors shaping US-China relations.

liazhu@chinadailyusa.com