Washington, D.C. – TechNet, the national, bipartisan network of innovation economy CEOs and senior executives, in a letter to Secretary of Commerce Howard Lutnick, urged the U.S. Department of Commerce to review the former Biden administration’s AI diffusion rule, highlighting concerns that overly broad export controls on AI chips and computing systems undermine U.S. innovation, weaken AI deployment, and cede critical digital infrastructure markets to foreign competitors, including China.

“AI chips and computing systems are critical not only to national security but also to global economic growth and U.S. leadership in innovation,” Linda Moore, TechNet President and CEO, states in the letter. “We support protecting U.S. national security and preventing advanced technologies from falling into adversarial hands.  However, smarter, more targeted export controls can achieve these goals without undermining American innovation and AI leadership.” 

The letter focuses on 3 priorities:

  • Targeted controls and enhanced monitoring mechanisms
  • Adopt diffusion strategy in collaboration with allies
  • Invest more in U.S. AI innovation and domestic capacity

Read the full text of the letter below. 

Dear Secretary Lutnick: 

AI chips and computing systems are critical not only to national security but also to global economic growth and U.S. leadership in innovation.  Export controls play a vital role in keeping technology out of the wrong hands, and we encourage the administration to review the Biden administration’s AI diffusion rule to ensure that the United States strikes the right balance between national security concerns and innovation. We believe the rule could be improved to better support U.S. innovation, enhance AI deployment capabilities, and maintain our competitive edge in critical digital infrastructure markets where global leadership is essential. 

The rule imposes limits on exporting advanced computing chips, creating restrictions that could be challenging for most countries.  This could significantly shrink the global market for American-designed chips.  Middle-tier nations, as outlined in the rule, will find that U.S. sales caps don’t meet their computing needs, pushing them to develop independent supply chains and seek alternative suppliers.  Countries such as Brazil, India, Israel, Iceland, Saudi Arabia, and the UAE — classified as middle-tier under the rule — have both the resources and the motivation to invest in computing ecosystems outside U.S. influence.  As a result, China and other restricted nations could source these technologies from non-U.S. producers, diminishing the rule’s effectiveness.  Instead, we encourage the administration to consider enhanced government-to-government security agreements that allow countries more compute access, keeping American companies competitive while ensuring security concerns are addressed. 

Export controls on semiconductors should not be a unilateral effort.  The administration must work with a broader group of allies beyond the tier-one countries designated in the rule.  A coordinated strategy for AI regulation and chip export controls is essential to securing semiconductor supply chains and maintaining national security.  The challenges of compliance gaps between the U.S. and its allies during previous restrictions highlight the risks of acting alone.    

Additionally, enforcing the new regulations would require exporters to aggregate global sales data, including from foreign competitors — a difficult and costly endeavor.  Combined with new, untested licensing requirements, these burdens could disrupt U.S. companies’ expansion plans in key markets.

We support protecting U.S. national security and preventing advanced technologies from falling into adversarial hands.  However, smarter, more targeted export controls can achieve these goals without undermining American innovation and AI leadership.  To strengthen U.S. policy, we recommend:

Targeted controls and enhanced monitoring mechanisms: 

  • Focus on targeted export controls that address national security threats and are based on specific risks rather than broad-based restrictions. 
  • Remove the regulatory distinction between open and closed source models. 
  •  Tighten loopholes in existing export controls and enhance verification systems already in place for better export control enforcement, including mechanisms to monitor the global flow of advanced chips to mitigate the risk of critical technology falling into the hands of U.S. adversaries. 
  • Ensure BIS is properly resourced to ensure that it can effectively enforce targeted export controls and provide assistance to industry as they seek to comply with their implementation. 
  • Move forward with some of the positive aspects of the AI diffusion rule, including a country approach rather than end-use approach, updates to the Notified Advanced Computing (NAC) and Advanced Computing Authorized (ACA) license exceptions, implementation of license exceptions AIA and ACM, removal of regulatory hurdles for authorized and allied countries, and more flexibility on small purchases.

Adopt diffusion strategy in collaboration with allies: 

  • Create a more dynamic and permissive diffusion framework for working with allies that ensures U.S. competitiveness.  The U.S. should encourage U.S. AI investment abroad and work with partners to secure semiconductor supply chains in support of an AI ecosystem that is independent, pro-innovation, and market-driven.
  •  Set clear and easily identifiable security standards for access to increased U.S. compute and the deployment of U.S. technology, rather than setting arbitrary computing power thresholds. 
  • Better harmonize export control regimes with allies to avoid compliance gaps. 
  •  Ensure compliance burdens for regulations remain lower than the cost of alternative AI stacks that are anchored outside the U.S. or partner countries. 
  • Consider limiting the output of China’s development by turning to the first Trump administration’s Huawei sanctions model that successfully stopped China’s rapid ascent in global telecommunications.  The use of Chinese AI technology in or by other countries is inevitable and already occurring.  Disincentivizing other countries from using or permitting the use of Chinese AI, particularly within their own governments, by offering American alternatives will help limit the proliferation of Chinese technology.

 Invest more in U.S. AI innovation and domestic capacity:

  • Continue substantial and sustained investments in U.S. innovation and domestic semiconductor chip manufacturing, including through easing permitting requirements and increased incentives. 

For America to maintain global AI leadership and prevent China from shaping the future of AI, we need an export control strategy that protects national security while fostering industry-led growth.  A restrictive, risk-focused approach will only push innovation and key markets toward China.  Instead, a policy that prioritizes targeted restrictions, strong enforcement, and strategic alliances will ensure the U.S. remains at the forefront of AI and computing innovation — keeping China from dominating the next wave of technological advancement.

Sincerely, 

Linda Moore

President and CEO