Advanced AI Models in National Security: Workshop Session

Published Date : 30/07/2025 

While large language models (LLMs) dominate headlines, a broader class of artificial intelligence systems are quietly transforming defense capabilities. Join the National Academies for a workshop session that delves into the evolving landscape of AI beyond LLMs, with a focus on national security. 

While large language models (LLMs) dominate headlines, a broader class of artificial intelligence systems are quietly transforming defense capabilities across domains from battlefield awareness to electronic warfare and autonomous systems.

Join the National Academies for a workshop session that will explore the rapidly evolving landscape of artificial intelligence models beyond large language models. The focus will be on the importance of model selection and design in the context of national security. Participants will engage with emerging approaches such as diffusion models for generative tasks, embodied AI for physical and autonomous systems, and AI techniques tailored for non-tokenizable data types including images, signals, and multidimensional sensor inputs.

The session will also examine the challenge of data sparsity, the role of simulated and synthetic data in accelerating development when real-world data is limited or classified, and the importance of selecting, training, and validating models appropriately for defense-relevant contexts. In doing so, the workshop will identify key areas where collaboration between academia, industry, and the Department of Defense can accelerate innovation.

Note: Registration is required to attend this event online. Please use the link above to register. Limited seats are available for in-person attendance at the National Academy of Sciences Building in Washington, D.C. If you’d like to join us in person, please email Gabrielle Risica at grisica@nas.edu. 

Frequently Asked Questions (FAQS): 

Q: What is the focus of the workshop session?

A: The workshop session focuses on exploring the evolving landscape of artificial intelligence models beyond large language models, with an emphasis on model selection and design in the context of national security.

Q: What types of AI models will be discussed?

A: The session will discuss emerging approaches such as diffusion models for generative tasks, embodied AI for physical and autonomous systems, and AI techniques tailored for non-tokenizable data types including images, signals, and multidimensional sensor inputs.

Q: How does the workshop address data sparsity?

A: The workshop will examine the challenge of data sparsity and the role of simulated and synthetic data in accelerating development when real-world data is limited or classified.

Q: Why is collaboration between academia, industry, and the Department of Defense important?

A: Collaboration between these sectors is crucial for accelerating innovation and ensuring that AI models are selected, trained, and validated appropriately for defense-relevant contexts.

Q: How can I register for the workshop?

A: Registration is required to attend the event online. You can register using the provided link. For in-person attendance at the National Academy of Sciences Building in Washington, D.C., please email Gabrielle Risica at grisica@nas.edu. 

More Related Topics :