type: timeline_event
On March 21, 2026, Deputy Secretary of Defense Steve Feinberg signed a memorandum designating Palantir's Maven Smart System as an official Department of Defense "program of record" — a bureaucratic milestone that locked in long-term stable funding and institutional permanence for what had become the Pentagon's primary AI-powered intelligence and targeting platform. The designation transferred primary oversight responsibility from the National Geospatial-Intelligence Agency (NGA) to the Chief Digital and Artificial Intelligence Office (CDAO), consolidating control over the program within the office that had been most aggressive in pursuing the Anthropic blacklist. Palantir's stock surged on the news.
Program of record status represented the most significant institutional commitment the Pentagon could make to a weapons or intelligence system short of a congressional appropriation. It meant Maven would be included in future budget submissions as a standing program rather than a pilot or experimental initiative, ensuring years of guaranteed funding and making it bureaucratically difficult for future administrations to cancel. Bloomberg reported that the designation was expected to translate into billions of dollars in contracts over the program's lifecycle, cementing Palantir's position as the dominant AI platform for military intelligence and targeting.
The designation created a glaring contradiction at the heart of the Pentagon's AI strategy. Maven's most sophisticated workflows ran on Anthropic's Claude, the very AI system the Pentagon had just blacklisted. Custom integrations between Maven and Claude Code — Anthropic's programming-focused AI tool — powered critical targeting, intelligence analysis, and operational planning capabilities that Palantir had spent months building and certifying for classified environments. By locking in Maven as a permanent program while simultaneously banning its underlying AI engine, the Pentagon had created a situation in which its flagship AI system depended on technology it had declared a supply chain risk. The contradiction underscored the degree to which the blacklist was driven by political rather than operational considerations, and it intensified pressure on the Pentagon to either resolve the Anthropic dispute or dramatically accelerate the transition to alternative AI models.