Positioned to Advance Key Software Technologies for National Security
U.S. private industry leads the world in artificial intelligence research, development, and deployment, having invested $109 billion in 2024 alone. AI attracts massive RD&D investment because the computational power to train models doubles every few months, and dataset sizes double about twice a year. Those and other circumstances lead to advances in large language model capabilities, deployment of agentic AI, continued evolution of robotics, and other results.
Further, in late 2025, the White House launched the Genesis Mission to accelerate AI development and use by combining private industry investment, the national laboratories’ supercomputing resources, federal data, and research at leading universities such as Carnegie Mellon University.
The SEI is an effective and efficient conduit to elevate invention by CMU researchers into results relevant to DoW software goals and to transfer national security and defense perspectives into future CMU basic research.
SEI Director and CEO
The CMU Software Engineering Institute, a federally funded research and development center, is in a key position in the AI RD&D landscape. We interact with private software industry firms, are sponsored by the U.S. Department of War, perform work plans with federal agencies, and offer an unparalleled connection with the leading AI research university in the nation, CMU.
The SEI is an effective and efficient conduit to elevate invention by CMU researchers into results relevant to DoW software goals and to transfer national security and defense perspectives into future CMU basic research. Further, as DoW’s only FFRDC chartered to look at software issues, the SEI is an objective, independent, trusted agent that can to work with software firms, tailoring new software practices they offer to DoW contexts. to tailor their new software practices for use in DoW contexts.
Currently, the SEI is collaborating with Accenture, a global professional services company, to develop, pilot, deploy, and transition an AI Adoption Maturity Model, with a specific focus on generative and agentic AI. We are an ideal fit for this work because of our expertise in maturity modeling and in applied software engineering, cybersecurity, and AI RD&D.
Recently, the SEI, working with CMU, established the new area of AI security, with the first-of-its-kind AI Security Incident Response Team (AISIRT). Because greater AI use introduced new varieties of threats, the SEI extended the CERT Coordination Center’s decades-long leadership in incident handling to AI. The AISIRT leads the response to AI security threats and incidents by being the trusted broker that brings together the stakeholders, such as the DoW, federal agencies, and commercial firms.
Making software technology a strategic advantage for national security is reflected, too, in other forward-looking SEI RD&D. Our work is, for instance, leading the advancement of a discipline for AI Engineering, exploring the intersection of cybersecurity and quantum computing, advancing robotics for manufacturing, and seeking ways to reduce the cost of using low-SWAP-C (size, weight, power, and cost) hardware.
As you will discover in this annual review, the SEI delivers guidance for the test, evaluation, verification, and validation of AI for lethal autonomous weapons systems; provides tools to help Air Force managers identify and evaluate AI and data science talent; builds cyber resilience for allies; and enables development of air combat capabilities for unmanned aerial vehicles, among other improvements for national security that rely on quality software.
The SEI’s capability to conduct leading-edge RD&D is nothing new. In 2025, we celebrated our 40th anniversary of enabling our national security organizations to deliver critical software capabilities when and where needed.
Paul Nielsen
