US Export Rule Hits AI Firms
The U.S. Commerce Department has introduced a new export control regulation, commonly referred to as the “50% rule,” which is expected to significantly affect AI and technology companies operating on a global scale.
The U.S. Commerce Department has introduced a new export control regulation, commonly referred to as the “50% rule,” which is expected to significantly affect AI and technology companies operating on a global scale. The rule, issued by the Bureau of Industry and Security (BIS), extends licensing requirements to subsidiaries that are more than 50% owned by entities listed on U.S. export control and sanctions lists. Officials say the move is designed to prevent sensitive technologies, particularly in artificial intelligence, from being indirectly transferred to countries deemed national security risks, with a particular focus on China (Axios).
Overview of the “50% Rule”
The “50% rule” changes the way U.S. export controls are applied to multinational companies. Previously, restrictions primarily targeted parent companies directly listed on control lists. Under the new rule, any subsidiary that is more than 50% owned by a listed entity now falls under the same licensing requirements. This means that even indirectly controlled subsidiaries are subject to U.S. export licensing and must obtain approval before transferring or sharing sensitive technology internationally.
Commerce Department officials argue that the rule closes a loophole in previous export controls that allowed sensitive technologies to be accessed indirectly through subsidiaries. According to a statement by BIS, “This adjustment ensures that U.S. controls reflect the realities of modern corporate ownership structures and the global reach of critical technology”.
Implications for AI Companies
The rule is likely to have widespread effects on AI firms with international operations. Companies may need to conduct thorough reviews of their ownership structures, partnerships, and supply chains to ensure compliance. Experts warn that this could result in increased administrative burdens, delays in licensing approvals, and potential disruptions to ongoing research and product development (Axios).
One example is Anthropic, a U.S.-based AI firm that has already implemented policies similar to the “50% rule.” The company reportedly severed partnerships with subsidiaries controlled by Chinese parent entities to ensure compliance. Analysts suggest that other AI companies could face similar strategic decisions as they navigate these regulations.
The rule also emphasizes the importance of legal and compliance departments within AI firms, as they must now closely monitor ownership changes, mergers, and acquisitions to avoid unintentional violations. Companies that fail to comply could face penalties, fines, or restrictions on exporting technology critical to AI development.
Industry Reactions
Industry responses to the new rule have been mixed. Some executives support the measure as a necessary step to protect national security, while others express concern that the regulation could stifle innovation.
Joseph Hoefer, principal and AI policy lead at Monument Advocacy, commented, “If American firms are tied up chasing ownership records and waiting for licenses to be approved, the U.S. risks slowing its own innovation while competitors in other regions advance unchecked.”
Conversely, advocates for the rule argue that controlling the flow of sensitive AI technology is essential to prevent strategic advantages from being transferred to adversarial nations. Kit Conklin, a former adviser to the House Select Committee on China, noted that the rule “addresses gaps in current export controls and ensures that subsidiaries cannot circumvent restrictions by operating under partial ownership structures.”
Balancing Security and Innovation
The introduction of the “50% rule” highlights the delicate balance between maintaining national security and promoting innovation. AI technology is increasingly critical for economic growth, defense applications, and commercial competitiveness. At the same time, sensitive AI capabilities, such as advanced machine learning models and large-scale data processing, could pose national security risks if transferred to foreign entities without proper oversight.
Some industry observers warn that while the rule strengthens national security, it may also discourage investment in AI startups and slow the deployment of cutting-edge technologies. Companies may hesitate to engage in international partnerships, fearing regulatory complexities and delays in approval.
Potential Global Effects
Because many AI firms operate in a multinational environment, the “50% rule” could have significant global repercussions. International subsidiaries of U.S. companies must now be evaluated under U.S. law, potentially impacting research collaborations, product development, and supply chain agreements abroad.
Experts suggest that AI companies may need to restructure ownership models or create separate legal entities to continue operations while complying with the new rule. Some smaller firms may find these adjustments particularly challenging due to limited legal and administrative resources.
The regulation may also encourage foreign AI companies to seek alternatives outside U.S. technology and expertise. By imposing stricter controls on U.S.-origin AI technology, competitors in Europe, Asia, or other regions might develop parallel systems, potentially reducing U.S. influence in the global AI ecosystem.
Policy Context
The “50% rule” is part of a broader effort by the U.S. government to control the international transfer of advanced technologies. It follows previous measures aimed at restricting semiconductor exports, surveillance tools, and other dual-use technologies. Policymakers argue that as AI becomes more pervasive and influential, controlling its distribution is vital to maintaining technological superiority in key sectors.
Commerce officials note that the rule applies specifically to technologies deemed sensitive, including AI tools that can be applied in defense, surveillance, or critical infrastructure. These measures are designed to prevent strategic technologies from benefiting nations with interests contrary to those of the United States.
Looking Forward
The AI industry now faces a period of adjustment. Companies must carefully evaluate ownership structures, partnerships, and licensing requirements to comply with the new regulation. At the same time, policymakers and industry leaders must consider ways to maintain the United States’ competitive edge in AI while ensuring security concerns are addressed.
Analysts expect that discussions between the U.S. government and industry stakeholders will continue in the coming months. Companies may seek clarifications, guidance, and potential exemptions for collaborative research projects. The final outcomes will likely shape how AI technologies are developed, shared, and deployed globally for years to come.