One of the more useful contributions of GAO’s April 2026 AI acquisitions report (GAO-26-107859) is its taxonomy of the different procurement approaches federal agencies are actually using—not as a policy prescription, but as an empirical account of what agencies have tried, what trade-offs they’ve encountered, and where each approach leaves agencies exposed.

Agency-Directed vs. Vendor-Driven

Some agencies began with a defined requirement and went out to acquire a solution. Others found vendors presenting AI capabilities to them that didn’t correspond to any existing requirement—and accepting those offerings anyway. GSA acquired a facility management software platform that included a chatbot feature the vendor added as a bonus, not in response to any stated requirement. VA awarded a task order for medical software that arrived with embedded AI capabilities.

Most of GAO’s April 2026 report on federal AI acquisitions (GAO-26-107859) documents failure modes—programs that didn’t document lessons learned, contracts that lacked AI-specific terms, programs retired without institutional postmortems. Two acquisitions stand apart as comparative benchmarks: DOD’s Maven program and GSA’s USAi platform. The report uses them to illustrate what AI acquisition looks like when agencies have had time to learn from their own mistakes.

Maven

Project Maven is DOD’s longest-running high-profile AI acquisition. Managed by the National Geospatial-Intelligence Agency, it uses machine learning and computer vision to analyze geospatial imagery and identify potential targets for human assessment. It has had a complicated history—including a period of public controversy over its AI ethics implications—but from an acquisition standpoint it represents an accumulation of hard-won institutional knowledge.

The Dark Eagle is not a single weapon so much as an integrated system of components developed across multiple contractors and shared, in part, with the U.S. Navy. Understanding how those pieces fit together clarifies both what the weapon can do and where its development risks have been concentrated.

The missile booster is a two-stage rocket developed by Lockheed Martin and Northrop Grumman. When mated with the hypersonic glide body, the complete assembly is designated the Navy-Army All Up Round plus Canister, or AUR+C. This combined form is what actually leaves the transporter erector launcher during a live-fire event. Critically, the same booster stack serves both the Army’s ground-launched LRHW and the Navy’s Conventional Prompt Strike system, which can be fired from surface vessels and submarines. That cross-service commonality is a deliberate acquisition choice — it spreads development cost and creates production efficiencies that neither service could achieve independently.

Military innovation rarely arrives as a single breakthrough. It tends to emerge as a convergence—of computation, doctrine, and necessity—until suddenly the battlefield looks fundamentally different. Autonomous drone swarms represent exactly that kind of shift. They are not merely an evolution of unmanned systems, but a redefinition of how force is applied, coordinated, and scaled.

Traditional drone warfare, as it developed over the past two decades, has largely been characterized by centralized control. Whether remotely piloted or semi-autonomous, drones have functioned as extensions of human operators—tools of precision, persistence, and surveillance. But this model carries inherent constraints: bandwidth limitations, operator fatigue, latency, and vulnerability to disruption. The introduction of swarm autonomy begins to dissolve those constraints.