National Security & Defense

National Security and International Affairs (NAICS 9281) covers the US Department of Defense (DoD), the intelligence community, and the State Department's international-affairs functions. The DoD is the largest US federal employer (~3.4 million people across active-duty military, reservists, and civilian DoD employees) with an annual budget exceeding $850 billion. The intelligence community includes the CIA, NSA, NRO (National Reconnaissance Office), DIA (Defense Intelligence Agency), and the broader Office of the Director of National Intelligence ecosystem — combined intelligence-budget exceeds $90 billion annually. Allied defense and intelligence services in the Five Eyes (UK, Canada, Australia, New Zealand) plus NATO partners account for similarly large investments. The defense-industrial base — Lockheed Martin, RTX (Raytheon), Northrop Grumman, General Dynamics, Boeing Defense, BAE Systems, plus the AI-defense newcomers (Anduril, Palantir, Shield AI, Saronic, Helsing) — supplies the technology platforms.

AI is reshaping every domain of national security. Joint All-Domain Command and Control (JADC2) initiatives integrate ML across air, sea, land, space, and cyber domains. Anduril Lattice and Shield AI Hivemind anchor the new generation of AI-native defense platforms competing with traditional prime-contractor offerings. Palantir Gotham (the dominant intelligence-analysis platform across DoD and the IC) and Palantir AIP integrate generative AI into operational workflows. Maven Smart System (originally Project Maven from 2017, now matured) applies computer vision to ISR (intelligence, surveillance, reconnaissance) data at scale. Autonomous and semi-autonomous platforms — drones (MQ-9 Reaper, Switchblade), maritime drones (Saronic, Anduril Dive-LD), ground unmanned systems — increasingly integrate AI for navigation, target-recognition, and mission planning. Cyber-defense AI (CrowdStrike Falcon, SentinelOne AI, Microsoft Defender) is heavily deployed across government networks. Generative AI is increasingly used for intelligence summarization, language translation (DARPA's automated-translation programs), and analyst-augmentation workflows. Frontier-model deployment on classified networks accelerated dramatically in 2026: by May 2026, the Department of Defense had awarded contracts to Google, SpaceX, OpenAI, NVIDIA, Microsoft, AWS, and Reflection AI to deploy AI on Impact Level 6 and Impact Level 7 classified systems — the most sensitive networks short of compartmented intelligence — explicitly framed as a vendor-diversification strategy rather than single-vendor reliance.

The May 2026 Pentagon classified-network procurement deal was a structural moment: contracts went to seven vendors (Google, SpaceX, OpenAI, NVIDIA, Microsoft, AWS, Reflection AI) on Impact Level 6 and 7 networks, but Anthropic was deliberately excluded. The exclusion stems from a contract-language standoff: Anthropic refused to permit Claude use for "all lawful purposes," arguing the clause could enable domestic mass surveillance or autonomous weapons. In March 2026 Defense Secretary Pete Hegseth designated Anthropic a "supply chain risk" — a tag historically reserved for foreign-adversary-linked vendors — and Anthropic filed suit to challenge the designation. Despite the freeze, the Pentagon is reportedly piloting Claude Mythos Preview for unreleased cybersecurity work, and the White House has reopened talks after CEO Dario Amodei met with Chief of Staff Susie Wiles. The episode is the first time a frontier US AI lab has chosen contract terms over defense access, and it surfaces a structural question for the next decade of AI procurement: how do governments source frontier models when the labs that build them have their own views on permissible use?

The defense and intelligence workforce is undergoing structural AI-driven transformation. Intelligence analysts increasingly work alongside AI summarization and entity-extraction tools that compress collection-to-analysis timelines from days to hours. Imagery and geospatial analysts are being heavily augmented by computer-vision AI (Maven, Palantir) — the human role is shifting toward exception-handling and strategic interpretation. Operations-research analysts, data scientists, and AI engineers inside the DoD and IC are growing roles. Cyber-defense roles are growing in absolute headcount despite AI-augmentation pressure. Acquisitions and program-management roles face elevated importance as the Pentagon navigates the shift from traditional prime-contractor systems to AI-native software platforms. Special-operations forces are integrating AI mission-planning and ISR tools into operational workflows. New roles: AI-policy and ethics officer at major commands, AI-acquisition program manager, autonomous-systems operator, JADC2 integration architect.