U.S. intelligence agencies are moving toward agentic AI systems but are taking a deliberate, staged approach rather than rushing deployment.
The Defense Intelligence Agency, National Geospatial-Intelligence Agency, and FBI described efforts to build underlying infrastructure, ensure data readiness, and maintain security and compliance safeguards before agents are used in operational missions.
While they have accelerated acquisition processes to move faster, agency leaders emphasized they are still in early planning stages and cannot skip steps, reflecting a recognition that premature deployment of autonomous AI systems could create significant risks.
What happened
Defense Intelligence Agency, National Geospatial-Intelligence Agency, and FBI officials said at the DoDIIS summit this week that their agencies are moving deliberately toward agentic AI (AI systems that act autonomously on tasks). DIA is running a 90-day sprint to build its first enterprise AI platform and adopting Model Context Protocol, an open-source standard for connecting AI to external systems. DIA is retooling its ChatDIA capability to support agents. NGA established an AI and Data Return on Investment Tracking task force. FBI reported 139 AI use cases overseen by an AI Review Board.
Why it matters
Intelligence agencies see agentic AI as essential for moving beyond chatbots, but they face significant hurdles: acquisition challenges, security risks, and the need to balance speed with thoroughness. DIA emphasized it is "in the very early stages" and cannot "just jump and deploy into agents" without building underlying infrastructure first. NGA highlighted that data readiness is critical—"data and AI are inextricably linked." FBI officials acknowledged they must test heavily with human backstopping before deploying AI in operational use. These deliberate timelines suggest the sector recognizes that premature deployment could create compliance, security, and trust problems.
What to watch
DIA has accelerated its acquisition process using other transaction agreements, completing eight or nine in weeks or days rather than months or years. FBI is using broad agency announcements and cooperative research and development agreements to connect with technology experts and make faster, more informed contracting decisions. Both approaches signal intent to speed deployment while maintaining oversight—success will depend on whether this balance holds as agencies move from pilots to mission-critical use.
Ask the AI about this article →
Intelligence agencies recognize agentic AI as a strategic capability and are actively moving toward it, but the speaker statements reveal a tension between urgency and caution. DIA's 90-day sprint to build an enterprise AI platform and adoption of Model Context Protocol signal serious intent, yet Kinney's repeated emphasis that "we're in the very early stages" and "we're not there yet with agents" underscores how nascent this transition remains. The staged approach—pairing agents with legacy data before mission use cases—reflects awareness that autonomous systems operating in classified, high-stakes environments carry irreducible risks: compliance violations, security breaches, or unpredictable agent-to-agent interactions could have national security consequences. NGA's establishment of an AI and Data Return on Investment Tracking task force points to a secondary concern: AI systems are only as good as their data inputs, and ensuring data readiness across legacy intelligence infrastructure is a longer, less visible problem than building the AI layer itself.
All three agencies have begun using non-traditional acquisition pathways—other transaction agreements, cooperative research and development agreements, and broad agency announcements—to compress timelines from years to weeks. This suggests the agencies are not waiting for a perfect regulatory or technical framework; they are acquiring and testing in parallel. However, the FBI's emphasis on an AI Review Board vetting all non-routine applications, case-by-case risk assessment, and "backstopping or paralleling with a human process" reflects the standard operating model for now: AI as a tool whose outputs must be verified before operational use, not as an independent agent.
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytime
Ask AI anything about this article. Q&As are published on this page for other readers too.
OpenAI's GPT-5.6 Sol, launched July 9, drove a 35 percent revenue increase this quarter, with enterprise reven…
OpenAI is previewing transparent background support for GPT-Image-2 through its API, allowing users to generat…
HP Korea has formed a partnership with Upstage, a large language model (LLM) startup, to advance its localized…
At the "AI on Chips: Semiconductor Industry Trends Forum" hosted by DIGITIMES, industry experts highlighted th…
Japan's Central Council for Education was asked Friday by education minister Yohei Matsumoto to develop recomm…
Stripe is in acquisition talks to buy AI startup OpenRouter for more than US$7 billion, according to Bloomberg…
