Within the architecture of modern warfare budgets there is a quiet physical logic: money begets prototypes, prototypes beget deployment plans, deployment plans beget doctrine. In FY2024 the Department of Defense doubled down on that mechanics of momentum for autonomy and related fields. The President’s FY2024 DoD request framed RDT&E as a central lever, with a record RDT&E topline and explicit line items for artificial intelligence and joint command-and-control programs that together create the fiscal traction necessary for widespread adoption of autonomous capabilities.
This is not merely about new toys. The budget language and programmatic pushes show an intent to make autonomy integral to every domain. The FY2024 request highlighted a $145 billion RDT&E investment and an artificial intelligence figure called out inside it. Those allocations are matched by dedicated funding streams for Joint All-Domain Command and Control, rapid experimentation, and office-level initiatives intended to accelerate transition from laboratory algorithms to operational systems. The result is a pipeline: experimentation funds lower near-term risk, program dollars de-risk production, and procurement buys scale.
Congressional activity in 2023 underwrote that pipeline rather than simply trimming it. As NDAA committees and appropriations drafts moved through the summer, lawmakers explicitly authorized and in some cases increased funding and authorities for the Department’s AI and autonomy efforts, and pushed for centralized platforms and clearer classification frameworks for autonomy. Committee text and trackers show Congress actively shaping how DoD buys and governs autonomy, including expanding institutional tools available to the Chief Digital and Artificial Intelligence Office to shepherd adoption. These legislative actions create a political momentum that complements the Pentagon’s budgetary momentum.
The upshot is a practical amplification effect that I term replicator momentum. Small, targeted investments in perception algorithms, autonomy middleware, and data infrastructure do more than produce prototypes. They replicate capability across programs because the same software stacks, sensors, and data-management patterns are portable. When a service can plug the same autonomy middleware into an aircraft, a surface vessel, and a ground vehicle, the marginal cost of fielding autonomy falls and adoption accelerates. That economic logic helps explain why modest RDT&E items in a single fiscal year can produce outsized changes in fielded capability over a five-year horizon.
But momentum is not destiny. Three structural gaps threaten to make this fiscal replication ethically and operationally reckless. First, acquisition and governance guidance lag. The Government Accountability Office warned in mid-2023 that DoD had not yet developed department-wide guidance to inform AI acquisitions, a gap that risks inconsistent requirements for testing, data rights, and assurance across programs. Without common acquisition rules of the road, replicated software can spread replicated weaknesses.
Second, institutional fragmentation complicates accountability. The budget erects new enterprise-level authorities and rapid prototyping lines, while services and program offices retain control over platform-level command and control. Congressional drafts in 2023 sought studies and frameworks to rationalize that split, but the tension remains: autonomy is being pursued both as an enterprise software problem and as a hardware program problem, and neither model alone suffices to ensure safe, auditable deployment. The danger is that autonomy ends up simultaneously everywhere and nowhere within the formal chains of responsibility.
Third, the ethics and legal frameworks have not been synchronized with the fiscal momentum. Funding for AI and autonomy can be read as a policy preference for speed and iterated advantage. Yet accelerating procurement without equally forceful investments in testing regimes, red-team vulnerabilities, and provenance for training data invites brittle systems into contested environments. The FY2024 request included explicit funds to accelerate experimentation and to tie digital initiatives to JADC2 efforts, but money alone cannot substitute for disciplined processes that ground technical capability in legal and moral accountability.
If we accept that budgets create technological trajectories, then the current fiscal architecture pushes the United States toward an ecosystem of proliferated autonomy: more drones, more collaborative platforms, more autonomy middleware reused across systems. That can be stabilizing if it is paired with governance that prevents runaway delegation of lethal choice, with procurement rules that preserve human-in-the-loop constraints where required, and with interoperable testbeds that measure failure modes under realistic stressors. The choices made in appropriation language and committee reports thus matter as much as the RDT&E numbers printed in the budget book.
Policy prescriptions follow from this diagnosis. First, Congress should condition transitions from prototype to procurement on demonstrable, standardized assurance metrics for autonomy performance and safety. Second, DoD must deliver department-level acquisition guidance for AI and autonomy that specifies data rights, lifecycle testing, and red-team obligations. Third, the services should be required to separate autonomy software chains of custody from individual platform vendors to prevent lock-in and to make independent validation tractable. Finally, increased funding for experimentation must be matched by resourcing the nontechnical oversight functions — legal, ethical, and human factors research — that make technical gains responsible. These steps align incentives so that replicator momentum becomes a force for orderly, auditable capability rather than an engine of accidental escalation.
The FY2024 budget era represents an inflection point. The United States can harness replicator momentum to renew deterrent advantage in an era of rising peer competition, but only if money and oversight travel together. Without that coupling, budgetary acceleration will simply manufacture new classes of problems: interoperable vulnerabilities, fractured accountability, and ethical drift. The task for scholars, engineers, and policymakers is to make sure our fiscal choices produce not only capability but also a durable architecture of responsibility.