By Vannevar LabsLearning Objectives
- Understand how AI accelerates open-source intelligence (OSINT) analysis for national security
- Evaluate the role of human analysts in reviewing AI-generated threat assessments
- Assess the surveillance, privacy, and civil-liberties questions that AI targeting of foreign actors raises
What Is Decrypt?
Decrypt is the flagship intelligence platform from Vannevar Labs, a defense technology company founded in 2019 by Brett Granberg and Nini Moorhead. Decrypt uses AI to ingest and analyze enormous volumes of multilingual open-source and non-traditional data — helping military and foreign-policy users understand, track, and respond to the activities of foreign actors.
The platform is applied to missions such as counter-disinformation, maritime domain awareness, and Indo-Pacific operations. The core value proposition is scale: AI can surface patterns across far more foreign-language sources than a human team could read directly, then hand structured leads to analysts. Human intelligence professionals remain responsible for judging, contextualizing, and acting on what the system surfaces.
💡Key Concept
Open-Source Intelligence (OSINT): Intelligence gathered from publicly available sources — news, social media, public records, shipping data, and other non-classified feeds. AI-powered OSINT tools translate, filter, and correlate these sources at machine speed so that human analysts can focus on interpretation and decisions rather than manual reading and translation.
Key Capabilities
- Multilingual analysis — ingests and processes foreign-language open-source data at scale
- Foreign-threat tracking — helps users understand and respond to the actions of foreign actors
- Counter-disinformation — surfaces and analyzes influence and disinformation activity
- Maritime domain awareness — supports monitoring of activity across contested waters
- Indo-Pacific focus — tailored to a priority theater for US defense and foreign policy
- Analyst workflow — structures raw data into leads for human intelligence professionals
⚠️Warning
AI targeting of "foreign actors" from open-source data raises real surveillance, privacy, and civil-liberties questions — especially around how outputs feed downstream kinetic or influence operations. Natural-language processing and machine-translation errors or bias can misidentify threats, so human review is essential before any assessment drives action. Because much of this work occurs in classified settings, effectiveness claims are largely company- and customer-asserted and difficult to independently audit.
Company Details
| Detail | Info |
|---|---|
| Company | Vannevar Labs |
| Founded | 2019 (Brett Granberg, Nini Moorhead) |
| Headquarters | Palo Alto, California |
| Employees | 200+ (December 2025) |
| Funding | ~$91.5 million raised over three rounds (as reported) |
| Valuation | ~$575 million (2024), following a $75 million Series B led by Felicis (January 2023) |
| Notable Contract | November 2024 production contract with a $99 million ceiling ($16 million obligated at award) |
| Website | vannevarlabs.com |
Key Takeaways
- Decrypt applies AI to analyze vast multilingual open-source data, helping defense users understand and respond to foreign actors across missions like counter-disinformation and maritime awareness
- Vannevar Labs, founded in 2019, has raised roughly $91.5 million and holds a November 2024 production contract with a $99 million ceiling
- The honest caveat: AI targeting of foreign actors carries surveillance and civil-liberties risk, translation and bias errors can misidentify threats, and classified-setting effectiveness claims are hard to independently verify
