The NGA is preparing a formal roadmap for artificial intelligence
The National Geospatial-Intelligence Agency is nearing release of a new framework for artificial intelligence that its director, Lt. Gen. Michelle Bredenkamp, described as the agency’s blueprint for becoming an “AI first organization.” In remarks at the GEOINT Symposium in Denver, Bredenkamp said the document is being finalized and will be made available soon.
The framework is intended to align with the Defense Department’s AI strategy while setting projects, lines of effort, and goals for the agency. According to Bredenkamp, it will cover operationalizing geospatial intelligence and AI across the intelligence cycle, modernizing business operations, revolutionizing acquisition, strengthening partnerships, and maturing AI governance.
The goal is speed, scale, and integration
The NGA occupies a central role in defense intelligence as the functional manager for geospatial intelligence, responsible for fusing, analyzing, and distributing data from government and commercial intelligence, surveillance, and reconnaissance satellites. In that context, AI is not being positioned as an isolated technical experiment. It is being framed as a way to handle volume, increase speed, and improve pattern recognition across large datasets.
Bredenkamp tied the AI push to a broader shift in how the agency sees itself. She said NGA’s vision extends beyond maps and images toward operating as a data agency that leverages multi-intelligence and artificial intelligence to produce geospatial intelligence superiority for decision-makers.
That framing matters because it suggests AI will be embedded not just in analytic tools, but in the agency’s institutional design, workforce planning, and procurement model.
Humans remain central to decision-making
Even as the agency pushes toward broader adoption, Bredenkamp emphasized that AI is not meant to replace human analysts. She said the NGA is building responsible AI systems in which humans remain in the loop on decisions that matter most.
Her formulation was direct: AI can handle volume, speed, and pattern recognition, but critical thinking, contextual understanding, and the ability to ask what information actually means remain irreplaceably human functions. The intended model is therefore not substitution, but what she called human-machine teaming excellence.
That is paired with a workforce agenda. Bredenkamp said the NGA wants every employee equipped with AI literacy tools and career pathways that make innovation normal across the agency. This suggests the blueprint will be as much about organizational adoption as technical deployment.
A Rapid Capabilities Office signals urgency
Bredenkamp also said the NGA has established a new Rapid Capabilities Office to speed the integration of innovative commercial technology. That move fits the broader message of urgency that ran through her speech. The agency is trying to implement a new vision at the pace of present and future threats, and it appears to view acquisition speed as a key bottleneck.
For defense agencies, that can be as significant as the AI models themselves. If commercial tools, data services, and software updates cannot be fielded quickly enough, an “AI first” strategy risks becoming a planning document rather than an operational shift. A rapid-capabilities structure is one way of signaling that the agency intends to shorten that gap.
From aspiration to doctrine
The most important development may be that the NGA is moving from talking about AI in broad aspirational terms to defining a concrete framework for how it will be used, governed, and integrated. Many defense organizations have embraced AI rhetoric. Fewer have articulated a detailed blueprint that spans operations, acquisition, partnerships, governance, and workforce readiness.
The success of the NGA’s effort will ultimately depend on execution. But the direction is now clearer. The agency wants AI operationalized across the intelligence cycle, embedded in the daily work of GEOINT, and governed in a way that preserves human judgment on consequential decisions.
That is a pragmatic defense vision for artificial intelligence: faster machine processing paired with human interpretation, delivered through a more agile institution. If the forthcoming blueprint turns those ideas into repeatable practice, it could become one of the more consequential AI implementation models inside the U.S. national security system.
This article is based on reporting by Breaking Defense. Read the original article.
Originally published on breakingdefense.com
