The US Critical and Emerging Technologies (CET) list has reduced its domain count from 18 to 14. But more significant than the number itself is the fact that the "National Security Science & Technology Strategy" (NSSTS), published in August 2026, has embedded the list into national security strategy and federal research and development (R&D) implementation principles. The 2024 version explicitly stated that the list should not be interpreted as a ranking of funding allocation or policy priorities. The new Appendix A now requires relevant agencies to focus R&D on maintaining US advantage in CET and to prioritize applications toward national security objectives.
The NSSTS is a 24-page document that fulfills the congressional requirement under 42 USC 19221 of the CHIPS and Science Act. Positioned to support the 2025 National Security Strategy, it no longer keeps the list of technology domains, the rationale for federal funding, and security-related demands as separate matters. Changes in the list's terminology are no longer merely an exercise in judging trends in product categories—they have become an exercise in choosing which security applications research resources should be tied to.
Reading the Reduction from 18 to 14 Domains: Consolidation vs. Word Deletion
The figure of 14 domains does not mean that four technology areas simply vanished. In the 2024 version, "Advanced Engineering Materials" and "Advanced Manufacturing" were merged into "Advanced manufacturing and materials" in the 2026 version. "Highly Automated, Autonomous, and Uncrewed Systems and Robotics" has also been absorbed into "AI and autonomy." These two reorganizations account for a reduction of two top-level domains.
The remaining reduction stems from the fact that the former "Advanced Gas Turbine Engine Technologies" and "Human-Machine Interfaces" no longer exist as top-level categories. However, this does not mean all components of the latter have disappeared entirely. Neurotechnology has moved to biotechnology, and brain-computer interfaces have been placed under future computing technologies. Augmented reality, virtual reality, and human-machine teaming no longer appear under the same names.
Meanwhile, there are clear substitutions among individual terms. "Advanced cloud services" and "high-performance data storage and data centers," which appeared under 2024's Advanced Computing, are absent from 2026's Future computing technologies. Advanced modeling and simulation and data processing/analysis technologies, which were in the same former domain, also cannot be found. In their place, the new list cites photonic, neuromorphic, and non-von Neumann computing architectures. It also gives examples of edge computing and devices for tactical environments, brain-computer interfaces, technologies enabling high-security computing, consumer security-hardened operating systems, and advanced spatial computing.
The boundaries have also shifted in energy. The 2024 category "Clean Energy Generation and Storage" listed a broad range of technologies spanning renewable energy to carbon management, including power generation, energy storage, and grid integration. The corresponding 2026 domain is nuclear energy, which enumerates advanced fission and fusion, space nuclear power and propulsion, and high-temperature/radiation-resistant materials. While it is true that the terms "battery storage" and "grid integration" have disappeared, this cannot be extended to a judgment that the importance of energy technology overall has declined.
AR and VR have also disappeared as terms from Appendix A. However, advanced spatial computing remains, and the NSSTS explicitly states that sub-domains are illustrative rather than exhaustive. The scope of the document is too narrow to read this as excluding all of AR/VR from critical national security technologies. Appendix A is a list selected for its relevance to national security—not an evaluation sheet for society or industry as a whole.
Added Terms Lean Toward Military Applications and Supply Chain Intersections
The 2026 version explicitly adds post-quantum cryptography and integrated photonics. The materials domain now includes 2D materials and high entropy alloys, along with digital thread and digital twins. The AI and autonomy domain now includes multi-agent systems, swarm intelligence, and autonomous agent identification/authentication. The new terminology is more concrete than a mere reshuffling of domain names. Cryptography, the integration of light and semiconductors, information connecting design through manufacturing, and the control and authentication of multiple agents have all come to the forefront as terms indicating connections to national security applications.
The treatment of AI has also changed. "AI assurance and evaluation technologies" and "technologies that improve AI safety, reliability, security, and responsible use," which appeared under the same names in the 2024 version, do not appear under those names in the 2026 version. Instead, "interpretability and control" and "robustness against adversarial attacks and AI security" are listed. This does not mean AI safety itself has been abandoned. Rather, the former comprehensive terms have been replaced with control and robustness directly tied to national security.
In semiconductors and microelectronics, design automation, manufacturing processes and equipment, beyond-CMOS, heterogeneous integration and advanced packaging, specialized hardware, and MEMS/NEMS continue to be included. 2D materials are newly cited as an example of new materials, and integrated photonics has also been added. In contrast, non-von Neumann architectures have moved to future computing technologies. This shows that manufacturing, materials, and computing architectures are no longer confined within a single domain but have been reorganized around which security applications they connect to.
Appendix B maps these technologies to battlefield advantage, homeland defense, and leadership in transformative emerging technologies. Battlefield advantage has three columns: a single column for "Space; Air; Long-range strike," the undersea domain, and national security AI and autonomy. On the homeland defense side, border security and nuclear deterrence/missile defense each occupy one column, for a total of four columns combined with cyber defense and biodefense. This structure foregrounds contributions to military advantage and homeland defense over the size of product markets. Supply chain resilience and protection of research outcomes are also placed as conditions within the main strategy for selecting and implementing CET.
Even Outside the List, Federal Support Does Not End
Even though the term "data center" has disappeared from CET, this does not mean federal support for computing infrastructure will disappear. The FY2028 R&D priorities memo NSTM-5/M-26-16 calls for federal computing infrastructure that can be differentiated from the private sector, such as high-security data centers for classified research and dedicated AI accelerators handling unique federal data. It calls for choosing public-private partnerships or procurement in cases where private computing resources can meet cost-effectiveness and research demand.
The same applies to battery storage. M-26-16 cites electrochemistry and solid-state ionics as fundamental chemistry and materials science, positing that these underpin energy generation and storage. The fact that the term "battery storage" has dropped from CET headings and whether related federal research becomes zero are separate matters.
The funding principle of the NSSTS is to concentrate resources where they can complement rather than compete with private investment. The government takes on areas where basic research or weak commercial profit incentives apply. For early- and mid-stage development, the strategy prioritizes public-private partnerships in Appendix A domains including AI, quantum, and nuclear energy, and cites biotechnology and space as examples. Late-stage development, in principle, leans more heavily on private sector dependence. M-26-16 also calls on each agency to increase the ratio of basic research relative to late-stage development, and requires agencies significantly expanding late-stage development to demonstrate why it could not be realized without federal support.
However, the document does not explain that data centers were removed from the list because of heavy private investment. The funding principle of role-sharing with the private sector and the deletion of terms cannot be definitively established as the same causal relationship. What can be confirmed is only that the list's terminology and funding allocation principles have been linked within the same national security strategy.
Legal Effects Are Not Determined by the List Alone
42 USC 19221(d) stipulates that the Director of OSTP, in consultation with OMB and others, shall ensure that recommendations and priorities for national security-related R&D contained in the latest strategy or reports are incorporated into the annual budget requests of federal research agencies. Accordingly, Appendix A becomes a strategic reference point connecting to budget requests. However, the list alone does not determine budget amounts, grant approvals, or export control classification numbers. What comes first is how it is reflected in each agency's FY2028 requests and technology-specific plans, and whether it secures congressional appropriations. Separate procedures based on respective laws and regulations are required if it is to influence export controls or investment review.
There is also no direct automatic change to what CFIUS (Committee on Foreign Investment in the United States) calls "critical technologies." CFIUS's definition is based on 31 CFR 800.215. It consists of items on the US Munitions List (USML) and certain Commerce Control List (CCL) items, along with nuclear-related items, select agents and toxins, and emerging and foundational technologies controlled under Section 1758 of the Export Control Reform Act. Listing or delisting in CET Appendix A does not immediately rewrite these legal definitions or export controls.
Evaluating this update requires more than just the number 14. One must observe how the ratio of basic research to late-stage development changes in the FY2028 budget request, which CET each agency links to security applications, and whether necessary regulatory revisions emerge. Only then will it become clear whether the terminology swap in Appendix A has actually reached the ground level of R&D and procurement.