The United States has announced a new 25% tariff on certain advanced artificial intelligence semiconductors, including Nvidia’s H200 AI chips destined for China. The decision marks a major escalation in Washington’s effort to reshape global semiconductor supply chains while balancing economic competition, national security, and the fast-moving AI race.
The move follows months of speculation about whether the administration of Donald Trump would impose tariffs on semiconductors. On Wednesday, the president signed a proclamation that formally introduced the tariff, targeting specific advanced chips produced outside the US but routed through the country before export.
What the New Tariff Covers
The 25% tariff applies to advanced AI semiconductors manufactured abroad that pass through the United States before shipment to customers in other countries. Among the affected products are Nvidia’s H200 chips and competing AI accelerators such as the AMD MI325X.
The tariff does not apply to all semiconductors. Instead, it focuses on a narrow category of high-performance AI chips that play a critical role in data centres, large-scale AI training, and advanced computing workloads.
Chips imported into the US for domestic use remain exempt. This includes semiconductors used for research, defense, and commercial purposes within the country.
Nvidia’s H200 Chips and the China Market
The decision formalises an earlier move by the US Department of Commerce, which allowed Nvidia to resume shipments of its H200 advanced AI chips to vetted customers in China in December.
Despite the new tariff, Nvidia welcomed the announcement. The company said the policy allows US chipmakers to compete globally while supporting American jobs and manufacturing.
Nvidia has reported strong demand for the H200 chips from Chinese firms. According to industry reports, the company had considered ramping up production due to a surge in early orders from China, highlighting the continued appetite for cutting-edge AI hardware despite regulatory hurdles.
Why the US Is Targeting Advanced AI Chips
The proclamation signed by President Trump highlights deep concerns about US reliance on foreign semiconductor supply chains. According to the document, the US currently manufactures only about 10% of the chips it needs.
The administration views this dependence as both an economic vulnerability and a national security risk. By imposing tariffs on certain foreign-produced chips, the White House aims to encourage domestic manufacturing and reduce exposure to geopolitical disruptions.
At the same time, the policy reflects a more selective approach than outright export bans. Rather than blocking shipments entirely, the US is allowing sales to approved Chinese customers, while adding cost pressure through tariffs.
Impact on Other Chipmakers
The tariff does not affect Nvidia alone. Chips from other companies, including AMD, also fall under the new rules.
This broader scope signals that the US is not targeting a single firm, but rather a class of advanced AI semiconductors. Analysts say this could influence pricing, supply decisions, and investment strategies across the global chip industry.
For US-based chip designers that rely on overseas fabrication, the policy introduces new complexity. Companies may need to rethink logistics routes, pricing models, or long-term manufacturing plans.
China’s Delicate Balancing Act
China faces its own strategic dilemma. Beijing wants to strengthen its domestic semiconductor industry and reduce reliance on foreign technology. At the same time, Chinese companies do not want to fall behind global rivals in AI development while waiting for local chipmakers to catch up.
According to reporting by Nikkei Asia, the Chinese central government is drafting new rules that would limit how many overseas semiconductors domestic firms can buy. These guidelines could still allow some imports of Nvidia’s AI chips, marking a potential shift from China’s earlier resistance to foreign chip purchases.
Such a move would reflect a pragmatic approach. China would continue building its domestic capabilities while selectively importing advanced chips to stay competitive in the global AI race.
Demand Remains Strong Despite Tariffs
Industry observers say the tariff may not significantly dampen demand for Nvidia’s H200 chips in the short term. Advanced AI semiconductors remain scarce, and alternatives often fail to match their performance.
For Chinese companies working on large-scale AI models, access to top-tier hardware remains critical. Many firms may choose to absorb the higher costs rather than slow their AI ambitions.
However, longer-term effects remain uncertain. Sustained tariffs could accelerate China’s push to develop domestic substitutes or encourage buyers to seek non-US routing options where possible.
Strategic and Global Implications
The tariff highlights the growing overlap between trade policy, national security, and artificial intelligence. Semiconductors now sit at the centre of geopolitical competition, with governments using tariffs, export controls, and industrial policy to shape outcomes.
For the US, the policy aims to protect domestic manufacturing while keeping American firms competitive abroad. For China, it adds pressure to advance its semiconductor self-sufficiency without stalling innovation.
As AI demand continues to surge worldwide, the effectiveness of such targeted tariffs will come under close scrutiny. Much will depend on how companies adapt, how regulators enforce the rules, and how global supply chains respond.
The coming months will reveal whether the tariff reshapes the AI chip market or simply adds another layer of cost to an already complex global technology race.
