The Trump administration's plan to expand American AI chip sales globally is hitting friction at the Bureau of Industry and Security, which is struggling with licensing delays, staff departures, and inconsistent policy enforcement.
President Donald Trump's strategy to increase global sales of US-made AI chips faces significant operational challenges within the federal agency responsible for export controls.
The Bureau of Industry and Security (BIS) is contending with three major obstacles that could slow the initiative:
Licensing Bottlenecks: The agency's approval process for chip exports is creating delays that frustrate chipmakers seeking to sell internationally. Complex licensing requirements are slowing transaction timelines, potentially allowing competitors to capture market share.
Staff Attrition: The BIS is experiencing personnel departures at a time when the agency needs expanded capacity to handle increased export applications. Fewer staff members mean slower processing and reduced capability to manage the workload.
Policy Inconsistencies: Unclear guidance from leadership has created confusion about which exports qualify for approval and under what conditions. The lack of consistent policy application has made it difficult for companies to plan exports or understand regulatory requirements.
These internal constraints directly conflict with the administration's stated goal of leveraging America's dominance in AI chip manufacturing as both an economic advantage and geopolitical tool. The US semiconductor industry has sought clearer, faster export pathways to compete more effectively in global markets while maintaining necessary national security controls.
The friction points highlight a broader tension: the administration wants to promote exports while maintaining strict controls on sensitive technology to adversaries. The BIS must balance these competing priorities while managing operational challenges.
Industry sources suggest that resolving these issues—through staffing increases, streamlined licensing processes, and clearer policy guidance—will be essential for the export push to succeed. Without improvements, delays could undermine the administration's economic and strategic objectives in the critical AI chip sector.
Amazon Web Services plans to deploy 2 million Nvidia Blackwell Ultra, Rubin, and Rubin Ultra GPUs across its data centers in 2027 and 2028, expanding its AI infrastructure beyond the 1 million GPUs announced in March.
Nvidia has agreed to acquire Hugging Face, the popular open-source AI model repository, for $12.9 billion. The deal consolidates Nvidia's control over critical infrastructure in the generative AI ecosystem.
Nvidia projects $108 billion in quarterly revenue within months, joining Amazon, Apple, and Alphabet as companies exceeding the $100 billion quarterly milestone. The chipmaker's latest earnings show record momentum driven by AI demand.
Nvidia forecasted third-quarter revenue of $108 billion, exceeding Wall Street's $104 billion estimate, while projecting fiscal 2028 revenue growth of approximately 70%. The stock jumped more than 4% in after-hours trading.