Original broadcast 6/22/25
Presented by Snowflake & Carahsoft
Artificial Intelligence is no longer just a concept or a future possibility at the Central Intelligence Agency—it’s operational, deployed, and actively reshaping how the CIA functions across mission and support activities. In a recent interview on Fed Gov Today with Francis Rose, CIA Chief Information Officer La’Naia Jones offered a deep dive into the agency’s AI journey, detailing how foundational infrastructure, strategic partnerships, and intentional experimentation have brought the technology into full deployment.
From Planning to Practice
Jones opened the conversation by confirming a major milestone: the CIA has moved beyond planning and pilot projects and is now implementing AI tools at scale. “We’re really using [AI] at all aspects of the agency,” she said, noting that use cases range from cybersecurity and finance to human resources. This whole-of-agency approach demonstrates that AI at the CIA is more than a mission enhancement—it’s becoming a strategic enabler of operational efficiency and innovation.
Rather than placing all emphasis on custom-built solutions, the agency has taken a hybrid approach. “We’re very fortunate to have vendor partnerships with cloud service providers and AI companies who have shared their models with us,” said Jones. These partnerships allow the CIA to blend proprietary intelligence data with external models, optimizing both internal innovation and industry-proven tools.
Choosing Between Off-the-Shelf and Proprietary Tools
Determining whether to use off-the-shelf or in-house AI solutions depends on several variables. “It could be the type of data. It could be the means or method that the data was obtained by,” Jones explained. CIA handles a wide array of intelligence types, from open-source intelligence (OSINT) to signals and cyber intelligence. This diversity necessitates a flexible technology portfolio. Jones stressed that when security and specificity are paramount, proprietary tools are often necessary, but commercial tools can streamline basic processes and accelerate delivery.
Speed, in fact, is a critical factor. “Speed could be the difference of a million things,” she said, underscoring that the faster intelligence can reach stakeholders, the more effective it becomes. In this respect, leveraging commercial AI helps reduce time-to-insight and gets actionable data to analysts and operators when it matters most.
The Cloud Foundation
Jones credited the CIA’s robust cloud infrastructure as a key enabler of its AI strategy. “Really, without the cloud foundation, we wouldn’t be able to make the AI models possible,” she said. The CIA’s transition to a commercial cloud enterprise began five years ago—just before the COVID-19 pandemic—and laid the groundwork for today’s rapid AI deployment.
The agency has worked closely with cloud service providers to build the compute power, storage capacity, and network resilience necessary for scalable AI. “The closer the data is to the customer, the more they can readily use it,” Jones explained, emphasizing the importance of data locality in maximizing value.
CIA’s early adoption of cloud also influenced the broader federal landscape. Jones noted that many civilian agencies were spurred into action by the CIA’s leadership, realizing that if the intelligence community could embrace cloud, so could they. “We wanted to make the data, the information, the intelligence, as quickly and rapidly available to our customers,” she said. That mindset has now extended to AI.
Building AI Maturity, One Use Case at a Time
Rather than tackling the hardest problems first, the CIA chose to build its AI maturity methodically. “We wanted to start with something small and then build out from there,” Jones said. Cybersecurity was a natural starting point, given its dynamic nature and the growing need for real-time threat detection. Working with external partners, the agency tested agentic models for threat monitoring and tipping, gradually expanding into other domains like human resources.
The COVID-19 era, with its surge in remote work and digital collaboration, also presented opportunities. Jones and her team identified use cases for digital connectedness, remote employee support, and HR process optimization using AI tools.
This iterative approach has paid off. “The models are available. We have a ton of use cases and partnerships, both internal and external,” said Jones. The CIA now collaborates closely with its various directorates and listens carefully to operational needs before deploying new tools. “We just take it one day at a time,” she added, echoing the measured confidence behind the agency’s digital transformation.
Leading by Example
As the federal government at large continues to wrestle with AI governance, infrastructure readiness, and mission integration, the CIA stands out as a model of how to do it right. The agency’s five-year effort to build the right foundation, cultivate vendor relationships, and foster internal capability has resulted in a program that not only accelerates intelligence but also enhances operational efficiency across the board.
Jones’s leadership reflects a pragmatic yet forward-thinking philosophy. The CIA’s success proves that when an agency invests in the right infrastructure, starts small with realistic goals, and fosters both internal and external collaboration, transformative technologies like AI can be safely and effectively scaled.
Key Takeaways:
-
CIA uses both proprietary and commercial AI tools depending on data sensitivity and use case.
-
A strong five-year cloud infrastructure journey underpins its AI deployment strategy.
-
Starting with small, manageable use cases enabled steady and scalable AI adoption across the agency.