U.S. Clarifies AI Chip Rules to Stop Indirect Exports to China 

The move reinforces restrictions on advanced semiconductor technology amid rising geopolitical competition in AI. 

The U.S. Department of Commerce has moved to tighten restrictions on the export of advanced artificial intelligence (AI) chips, closing what officials and industry experts viewed as a loophole that could have allowed Chinese companies to access cutting-edge semiconductor technology through overseas subsidiaries.

The Bureau of Industry and Security (BIS) issued guidance on Sunday clarifying that export license requirements apply not only to entities located in China but also to companies headquartered in China that operate through subsidiaries abroad. The move is aimed at preventing Chinese firms from obtaining advanced AI processors, including Nvidia’s latest Blackwell chips, through operations based in countries outside China.

The guidance comes amid concerns in Washington that some Chinese technology companies may have been acquiring restricted chips through subsidiaries located in markets such as Malaysia and other regions. A policy paper recently circulated among U.S. officials reportedly warned that existing controls left an opening that could be exploited to secure high-performance AI hardware despite broader export restrictions.

While the exact number of chips that may have been shipped through such channels remains unclear, industry sources suggest that significant volumes could have been involved over the past year. The issue gained attention after the Trump administration decided in May 2025 not to enforce the Biden-era AI Diffusion Rule, which had established licensing requirements governing global access to advanced AI chips.

- Advertisement -

According to BIS, the latest guidance does not introduce new export controls but instead clarifies regulations that have been in effect since 2023. The agency said it remains committed to enforcing export restrictions to safeguard critical American technologies and limit their use by strategic competitors.

Nvidia stated that the clarification does not affect its existing obligations, noting that the company already faces licensing requirements for shipments of advanced AI chips to restricted entities. However, some experts argue that challenges remain, particularly around ensuring that chip foundries and supply-chain partners can identify whether products are ultimately destined for Chinese-linked organizations through indirect channels.

- Advertisement -
Loading form…
Nikita Kumari
Nikita Kumari
Nikita Kumari is a Journalist at EFY. She decodes deals, investments, and policy shifts, redefining the semiconductor and tech landscape.

Industry's Buzz

Image for representation purpose

Hanwool Wins Delta Group Order For MLCC Inspection Equipment

0
Hanwool Semiconductor has secured an order from a Delta Electronics subsidiary for its locally developed MLCC inspection equipment, challenging Japanese dominance and targeting growing...
DRDO

DRDO To Lead Next-Generation Semiconductor Discussions At ESTIC-2026

0
DRDO will bring semiconductor industry leaders and researchers together at ESTIC-2026 to advance next-generation chips, AI hardware and India's domestic microelectronics capabilities. The Defence Research...
Huawei

Huawei Ships 100m Kirin Processors as Domestic Chips Advance

0
Chinese tech leader expands Kirin silicon footprint and tests next-generation processor amid US trade curbs.  Huawei Technologies has shipped more than 100 million devices powered...
Semiconductor Chip (Representation Image)

IIT Hyderabad, Japan’s Mie University Partner on Semiconductor Research

0
IIT Hyderabad and Japan’s Mie University have signed MoUs to strengthen academic exchanges, joint research, and talent development in semiconductors and digital technologies. The Indian...
ASMl Holdings

ASML Lifts Sales Outlook as AI Tool Demand Exceeds Supply

0
Semiconductor giant warns capacity shortages for chipmaking tool orders will extend to 2028.  ASML has raised its full-year sales forecast following strong demand for its...

Learn From Leaders

Does India’s Chip Packaging Growth Need Local Materials At Scale?

0
What will it take for specialised semiconductor materials manufacturing to grow in India? Packaging volumes will be a key factor as OSAT facilities move...
Ronak Kumar Samantray, Founder, TM2Space Technologies Pvt Ltd.

“Our goal is to build data centres in space.” – Ronak Kumar Samantray, TM2Space

0
A space startup aims to build orbital data centres, letting users run AI models and store data on satellites while also supplying satellite hardware...

India’s Photonics Puzzle Is Finally Coming Together 

0
Can India turn its scattered photonics capabilities into a stronger industry? A new association aims to connect the ecosystem and strengthen the domestic value...
Jin Bains, CEO, Mini-Circuits

Mini-Circuits CEO Jin Bains on Navigating High-Frequency Component SWaP-C Constraints, Emerging Markets, and Career...

0
In this exclusive, Jin Bains, Chief Executive Officer (CEO) of Mini-Circuits, explains in a conversation with Dilin Anand of EFY about areas of emerging...

India Needs Not a Hundred, But Thousands of Chip Diversified Companies 

0
Can India build a semiconductor ecosystem beyond fabs? Indiesemic believes thousands of diversified chip companies are needed to make that possible. India’s semiconductor push is...

Startups

Compcarta Reduces The Cost And Complexity Of End-To-End Chip Design

0
Building a chip should not require assembling an entire design ecosystem. A Noida-based startup is developing a browser-based EDA platform that aims to reduce...
PowerMon

Aquaexchange Brings Monitoring And Automation To Shrimp Farming

0
The startup has developed an AI- and IoT-enabled platform that helps shrimp farmers manage farms, access financing, monitor operations, and connect with markets. The startup...

Drone-Like Aircraft For Logistics From BluJ Aerospace

0
Hyderabad-based BluJ Aerospace is developing indigenous electric and hydrogen-powered vertical take-off and landing (VTOL) aircraft for remote cargo transport. BluJ Aerospace, a Hyderabad-based aerospace startup...
Jash Jetly, Co-founder and CEO of EVE

“The cloud isn’t the only way to run AI. We’re proving it can happen...

0
A startup is building a decentralised AI and communication network that runs across everyday devices instead of centralised cloud servers. In an interview, Jash...
Mohit Gupta, founder and CEO, TYLsemi

Building AI Chips, One Block at a Time 

0
What if building an AI chip became as simple as assembling reusable blocks? The startup is working to make custom silicon accessible beyond the...