Pentagon SBIR Solicits AI Decoy Personas for Cellphone Metadata Masking; Solicitation Opens Sept. 23

The Department of Defense opened a Small Business Innovation Research solicitation on September 4 seeking AI-generated "personas" to mask cellphone digital exhaust—location data, device IDs, and behavioral patterns—by generating realistic fake traffic that defeats automated network analysis. Phase I runs through October 21 and targets a sub-2-pound device with 36-hour battery life; Phase III deliverables include hardened decoy boxes sized as Wi-Fi routers and personal pucks.

The device weight and battery constraints (sub-2 pounds, 36-hour endurance) suggest a tactical-unit scale deployment—squad or company level—not strategic air-defense or carrier-group masking. If the Phase I winners can actually generate behavioral profiles that fool both rule-based and ML-based traffic analysis, the second-order effect is immediate: every U.S. forward-deployed unit in the Indo-Pacific gets a $X-per-unit defensive layer against PLA SIGINT collection, which reshapes the cost calculus of any PLA contingency operation that relies on pre-conflict pattern-of-life targeting. If the AI personas fail against state-grade adversaries, the program becomes a technology demonstration with no operational impact—and the Pentagon's traffic-analysis defense posture remains exposed.

WHY IT MATTERS

Adversaries currently use cell metadata to track military personnel, identify VIP movements, and map sensitive facilities; a working decoy system flips that intelligence advantage.

The SBIR timeline (Phase I concept, Phase II prototype, Phase III production) suggests operational deployment by 2028–2029, which means the Pentagon is betting on AI-generated behavioral mimicry as a counter to PLA and Russian signals intelligence collection—a capability gap that has no current fielded solution. Watch the Phase I contract awards in late October; the winning design will telegraph how the Pentagon expects to defeat pattern-of-life analysis at scale.

WHAT THIS DOESN’T TELL US

Does the solicitation specify which adversary AI models (PLA traffic-analysis systems, Russian SIGINT platforms) the decoy personas are designed to defeat, or is this a generic capability against any automated network analysis?

Fault Lines

Strategic intelligence, synthesized daily — with a public track record. Every call graded against what actually happened.

Front page → Get the weekly brief →