AI/ML Basic Research Hub
The Artificial Intelligence / Machine Learning (AI/ML) Basic Research Hub is a consortium of industry, government, and academia focused on AI basic research originating from world leaders in academic research pertaining to AI/ML breakthrough technologies for future application to Army-relevant areas such as object recognition using Multiple Cooperative Autonomous Sensors, leader decision-making, replication of tactical behaviors to enable autonomous capabilities for maneuver, predictive maintenance, Intel support for Operations, network and cybersecurity, AI-enhanced common operating picture, intelligent business and process automation, and medical support. Collaboration between academia, industry, and government is a key element of the Hub concept as each member brings with it a distinctly different approach to research. Academia is known for its cutting-edge innovation; the industrial partners are able to leverage existing research results for transition and to deal with technology bottlenecks; and Army AI researchers keep the program oriented toward solving complex Army technology problems. Work in this project compliments Program Element (PE) 0602180A (Artificial Intelligence Technologies) and PE 0603040A (Artificial Intelligence Advanced Technologies). Work in this project is performed by the Artificial Intelligence Integration Center (AI2C).
FY2026 to FY2027 change Funding decrease reflects the termination of this initiative and the strategic reallocation of resources to support the evolving priorities.
FY2026 plans — current year Will research techniques to extend foundational models (such as those for language, vision, and segmentation) across multiple modalities; expand new methods to synthesize multi-modal data for use-cases such as querying the data through natural language, question-answering, semantic segmentation, and product generation.
FY2025 accomplishments Research techniques to efficiently and accurately transfer foundation models to improve automated threat recognition. Expand on current research to improve methods for making robust predictions in domains with limited observations and labels. Develop new methods to synthesize multi-modal data for use-cases such as querying the data through natural language, question-answering, semantic segmentation, and product generation.
FY2026 to FY2027 change Funding decrease reflects the termination of this initiative and the strategic reallocation of resources to support the evolving priorities.
FY2026 plans — current year Will research methods for rapid training, retraining, deploying, governing, and interacting with machine learning models hosted on robotic platforms and edge devices; develop new methods for communicating with and between machine learning models and edge devices; expand research into deploying state-of-the-art models, including but not limited to models generally considered to be large or compute intensive, onto rugged edge hardware and small form factor devices; conduct foundation research into methods for attacking and compromising machine learning and artificial intelligence systems as well as for defending against similar attacks; expand research to AI-enabling computing infrastructure…
FY2025 accomplishments Research improvements to AI-enabling computing infrastructure, devices, and algorithms for both enterprise and tactical computing environments. Research autonomy for robotic systems and methods for training, deploying, retraining, and governing machine learning models hosted on robotic platforms and edge devices. Conduct foundation research into methods for attacking and compromising machine learning and artificial intelligence systems as well as for defending against similar attacks.
FY2026 to FY2027 change Funding decrease reflects the termination of this initiative and the strategic reallocation of resources to support the evolving priorities.
FY2026 plans — current year Will extend research on human and non-human behavior and interactions in various online social settings; research effective occupational training in artificial intelligence and machine learning for an audience with diverse technical skills to improve the Army's capability to deploy and use AI/ML products; expand research methods for making machine learning output more interpretable for human consumption and the effects these techniques have on human decision making; research the use of quantitative metrics in measuring the ethical compliance of AI systems; expand research in novel algorithms for improving human decision-making.
FY2025 accomplishments Research human and non-human behavior and interactions in various online social settings. Extend current research on effective occupational training in artificial intelligence and machine learning for an audience with diverse technical skills to improve the Army's capability to deploy and use AI/ML products. Research methods for making machine learning output more interpretable for human consumption and the effects these techniques have on human decision making.