What is the GAIN AI Act and Why Does Nvidia Oppose it?

The AI chip industry is bracing for another wave of export restrictions as US lawmakers push legislation requiring domestic companies to prioritise American customers over international buyers. 

The Guaranteeing Access and Innovation for National Artificial Intelligence Act of 2025 (GAIN AI Act), introduced as part of the National Defense Authorisation Act, would require American developers of AI processors to prioritise domestic orders for high-performance processors before supplying them to overseas buyers.

This development is unfolding as Washington grapples with balancing national security concerns against the commercial interests of US tech companies. 

The Trump administration recently rescinded the Biden-era AI Diffusion Rule, which had imposed similar restrictions on computing power exports through a complex country-tier system. 

That earlier framework aimed to “regulate the global diffusion” of advanced AI chips and models while keeping powerful AI systems within the US and close allies.

Understanding the GAIN AI Act framework

The GAIN AI Act is a more straightforward approach to chip export controls than its predecessor. 

The legislation would mandate that American semiconductor companies fulfil domestic orders before shipping advanced processors overseas, effectively creating a US first policy for AI chip distribution.

Under the proposed rules, exporters would need to obtain licences and approval for shipments of silicon exceeding certain performance thresholds. 

The act specifically targets chips with total processing power of 4,800 or above, establishing clear technical benchmarks that determine which processors require export authorisation. 

This is a change from the Biden administration’s complex multi-tier country system to a more binary approach focused on chip capabilities.

The legislation aims to address supply constraints that have left American companies competing with international buyers for access to cutting-edge AI hardware. 

By prioritising domestic customers, lawmakers hope to ensure that US businesses, research institutions and universities can secure the computing power needed for AI development without lengthy procurement delays.

Why Nvidia calls the restrictions unnecessary

Nvidia has pushed back hard against the proposed legislation. 

The company argues it already serves American customers first, with Reuters reporting that a spokesperson says: “We never deprive American customers in order to serve the rest of the world.

“In trying to solve a problem that does not exist, the proposed bill would restrict competition worldwide in any industry that uses mainstream computing chips.”

The graphics processing unit (GPU) manufacturer has drawn unfavourable comparisons between the GAIN AI Act and the now-rescinded AI Diffusion Rule.

“The AI Diffusion Rule was a self-defeating policy, based on doomer science fiction and should not be revived,” the company says.

“Our sales to customers worldwide do not deprive US customers of anything – and in fact expand the market for many US businesses and industries.

“The pundits feeding fake news to Congress about chip supply are attempting to overturn President Trump’s AI Action Plan and surrender America’s chance to lead in AI and computing worldwide.”

Company filings show that American customers already represent Nvidia’s largest market segment, accounting for 49.9% of revenue in fiscal 2024, compared to 28% from China and 18% from Singapore. 

The proposed restrictions would target chips with specific performance capabilities. 

The legislation calls for denying “licenses for the export of the most powerful AI chips, including such chips with total processing power of 4,800 or above.” 

These technical thresholds mirror previous export control frameworks and would affect various AI accelerators currently in production.

GPUs have become critical infrastructure for training large language models (LLMs)  and other AI systems requiring massive computational resources. 

The new rules would potentially apply even to older AI GPUs like Nvidia’s HGX H20 or L2 PCIe models, which were specifically designed to comply with previous export restrictions while serving international markets.

The industry split on domestic prioritisation

The legislation has divided industry opinion. 

Brad Carson, President of Americans for Responsible Innovation, supports the measure, arguing: “Globally, these chips are currently supply-constrained, which means that every advanced chip sold abroad is a chip the US can’t use to accelerate American R&D and economic growth.

“As we compete to lead on this dual-use technology, including the GAIN AI Act in the NDAA would be a major win for US economic competitiveness and national security.”

The supply constraint argument shows broader market dynamics where demand consistently exceeds manufacturing capacity. 

Data centre operators, cloud providers and research institutions compete globally for access to the latest semiconductor technologies.

The GAIN AI Act specifically aims to ensure American “small businesses, start-ups and universities” can access the latest AI GPUs before international clients in countries of concern.

Brad concludes: “American buyers should be prioritised ahead of others, according to US legislators.”