TL;DR: Shield AI is a San Diego defense-technology company building autonomous flight software (Hivemind) and an autonomous fixed-wing unmanned aircraft called V-BAT. Founded in 2015 by brothers Brandon and Ryan Tseng plus Andrew Reiter, the company closed a $1.5 billion Series G in March 2026 at a $12.7 billion post-money valuation — up 140% from $5.3 billion twelve months earlier. Projected 2026 revenue: $540+ million, with more than half coming from international customers. Hivemind has been demonstrated controlling F-16 fighters in DARPA dogfighting experiments and flying Anduril’s Fury aircraft.
Why read: Shield AI is one of the three most-important new defense companies alongside Anduril and Swift Beat. It occupies a different niche — autonomous-flight software — that matters across the entire next-generation military aviation roadmap.
Best for: Anyone tracking defense AI, private-market defense exposure, or autonomous-aircraft technology.
Skip if: You already follow defense procurement professionally. Daily AI fundamentals in our free Beginners in AI newsletter.
If Anduril is the most-discussed defense-tech company of the past decade by valuation and Swift Beat is the highest-volume drone producer, Shield AI is the autonomy-software specialist that pilots them. The company’s Hivemind autonomy stack is being designed to fly the next generation of US Air Force collaborative combat aircraft — the uncrewed jets the Air Force plans to deploy alongside the F-35 in the 2030s.
Here’s the factual story of who built Shield AI, what it makes, and where the $12.7 billion valuation comes from.
Who founded Shield AI?
Shield AI was founded in 2015 in San Diego by three co-founders:
- Brandon Tseng — CEO at founding, now President. Former Navy SEAL. The company’s origin story comes from his combat experience: he wanted a way to give small units the ability to send a drone into a building or compound autonomously, in environments where GPS and communications were unavailable. Brandon attended Harvard Business School after his Navy service.
- Ryan Tseng — Brandon’s older brother. Currently CEO. Background in engineering and entrepreneurship; previously founded WiPower, a wireless-charging company that was acquired by Qualcomm.
- Andrew Reiter — co-founder and head of autonomy engineering. Engineering background spanning robotics, autonomous-flight algorithms, and the broader autonomy stack that became Hivemind.
The founding logic was specific. The Tseng brothers and Reiter believed autonomous flight in GPS-denied, communication-denied environments was the hardest unsolved problem in defense robotics, and the team that solved it would have a durable advantage. Ten years later that thesis has held up.
What does Shield AI actually make?
Two flagship products and a small number of supporting platforms.
- Hivemind — the company’s autonomous-flight AI stack. The stack lets a drone or aircraft navigate, identify objects, plan routes, complete missions, and operate in coordination with other aircraft entirely without operator input or external infrastructure. Hivemind is what makes everything else Shield AI ships strategically meaningful.
- V-BAT — the company’s fixed-wing unmanned aircraft. V-BAT is a vertical-takeoff-and-landing (VTOL) tail-sitter design that doesn’t need a runway. It can fly long-range missions (8–13 hours depending on configuration), carry various payloads (ISR cameras, electronic-warfare modules, communications relays), and operate from forward bases, ships, or austere environments.
- Nova — an earlier small quadcopter for indoor/urban autonomy work. Originally targeted at the small-unit reconnaissance use case Brandon Tseng described from his SEAL service.
- Hivemind Enterprise — a developer platform allowing third parties to build autonomous applications on the Shield AI autonomy stack. Announced with the 2025 $240M raise at the prior $5.3B valuation.
V-BAT is the visible product in operational deployment. Hivemind is the long-term strategic moat — the autonomy stack the company expects to power next-generation collaborative combat aircraft and other autonomous platforms across the defense industry.
What is Hivemind capable of?
- Autonomous flight in GPS-denied environments. Visual-inertial navigation lets Hivemind-equipped aircraft fly indoors, underground, in heavy electronic-warfare environments, and around terrain features.
- Fighter-jet integration. Hivemind has been demonstrated controlling modified F-16 aircraft during DARPA’s autonomous-dogfighting experiments. The same software has flown Anduril’s Fury collaborative combat aircraft.
- Swarming and multi-aircraft coordination. Multiple Hivemind-controlled aircraft can operate as a coordinated unit without per-aircraft operators.
- Target recognition and classification. Computer-vision models trained on military target classes power identification within defined rules of engagement.
- Mission completion under degraded communications. If the link to the operator is lost, the aircraft continues the pre-defined mission rather than aborting.
The strategic implication: a US Air Force pilot in an F-35 in the 2030s is expected to fly with a wingman of 4–6 uncrewed Hivemind-controlled fighters. Shield AI is one of the leading candidates to provide that autonomy.
What is V-BAT?
V-BAT is Shield AI’s small-tactical fixed-wing unmanned aircraft. The platform was originally designed by Martin UAV; Shield AI acquired the company in 2021 and absorbed V-BAT into its lineup.
Key V-BAT characteristics:
- Tail-sitter VTOL design. Lifts off and lands vertically, eliminating the need for runways, catapults, or large recovery infrastructure.
- Long endurance. Flight times of 8 to 13 hours depending on payload and configuration.
- Modular payloads. Different mission kits for ISR (intelligence, surveillance, reconnaissance), electronic warfare, communications relay, or other roles.
- Compact footprint. Can launch from ship decks, forward operating bases, or austere ground locations.
- Hivemind autonomy. Operates with full autonomous capability including in GPS-denied or communications-denied environments.
V-BAT 5.3, the current operational variant, completed July 2025 testing with 100% performance against Key Performance Parameters — an unusual all-pass result for any defense test cycle.
What contracts is Shield AI executing?
| Customer | Program | Value |
|---|---|---|
| U.S. Coast Guard | V-BAT maritime ISR services IDIQ contract (IDIQ = Indefinite Delivery Indefinite Quantity, a federal-contracting structure that pre-approves spending up to a ceiling without committing to specific orders) | $198 million |
| Netherlands Ministry of Defence | V-BAT procurement for Navy/Marine Corps | 8 aircraft |
| Romania | V-BAT international sale | Undisclosed |
| Indonesia | V-BAT international sale | Undisclosed |
| Japan | V-BAT international sale | Undisclosed |
| U.S. Air Force, DARPA | Autonomous combat aircraft experiments (Hivemind) | R&D scale |
| Replicator Initiative (DoD) | V-BAT for operational deployment | Multi-aircraft awards |
The international mix is unusually high for a US defense-tech company at Shield AI’s stage — more than 50% of projected 2026 revenue is from non-US customers. That international diversification is part of why the Series G investors paid the 140% valuation step-up.
Funding history
| Round | Year | Valuation | Notable investors |
|---|---|---|---|
| Series A–D | 2016–2022 | Various; growing | Andreessen Horowitz, Riot Ventures, others |
| Series E | 2022 | $2.7B | L3Harris and existing investors |
| Series F | 2025 | $5.3B | L3Harris, Riot Ventures (Hivemind Enterprise focus) |
| Series G | March 2026 | $12.7B post-money | Advent International (lead), JPMorgan Chase Security and Resiliency Initiative (co-lead), Blackstone preferred |
The Series G raise was structured as $1.5 billion in primary equity plus $500 million in preferred equity from Blackstone plus a $250 million delayed-draw facility — roughly $2.25 billion in total capital available.
How does Shield AI compare to Anduril?
Both companies grew up in the same new-defense-tech ecosystem. They sometimes compete (notably for collaborative combat aircraft autonomy) and sometimes partner (Hivemind has flown Anduril’s Fury). The structural difference:
- Anduril is a broad-portfolio hardware-and-software company. Lattice OS + drones + counter-UAS + maritime + electronic warfare across one integrated stack.
- Shield AI is autonomy-software-first. Hivemind is the core; V-BAT is the platform that proves the autonomy. The bet is that autonomy software, not the airframe, is the durable strategic asset.
Both bets can be right. Anduril’s portfolio approach generates more total revenue. Shield AI’s autonomy specialization positions it to be the AI layer inside aircraft built by other companies — potentially including Anduril’s own future platforms.
What about the Brandon Tseng leadership transition?
Brandon Tseng was Shield AI’s founding CEO. The role transitioned to his brother Ryan Tseng as the company scaled, with Brandon shifting to President and remaining the primary public-facing voice. The arrangement parallels several other technology companies where the founder takes a customer-facing/strategic role while a co-founder or hired executive handles day-to-day operations.
Brandon’s SEAL background has been central to Shield AI’s product narrative since founding — the use cases he describes from his service are the use cases the autonomy stack was designed for.
FAQ
Is Shield AI publicly traded?
No. Shield AI is privately held. The Series G post-money valuation is $12.7 billion. An IPO has been discussed publicly but no formal filings have been made.
Who are Shield AI’s biggest customers?
The US Coast Guard (V-BAT maritime ISR), the broader Department of Defense (V-BAT for Replicator and adjacent programs), and international defense customers (Netherlands, Japan, Romania, Indonesia). More than half of projected 2026 revenue is international.
What is Hivemind Enterprise?
Hivemind Enterprise is the developer platform Shield AI announced with the 2025 Series F. It lets third parties (defense customers, partner integrators) build autonomous-flight applications using Shield AI’s autonomy stack as the foundation. The strategic logic is similar to Anduril’s Lattice OS — create an ecosystem around the software so that customers buying into it generate compounding value.
Does Shield AI build weapons?
V-BAT is an unmanned aircraft, not inherently a weapon. The platform supports payloads that include weapons (strike modules) as well as non-kinetic payloads (ISR, communications relay, electronic warfare). Hivemind is software that can be applied to autonomous combat aircraft. The company operates within the DoD’s rules of engagement and DoD Directive 3000.09 governing autonomy in weapon systems.
How does Shield AI relate to the F-35 program?
Shield AI does not build F-35-class aircraft. The relevant intersection is the Air Force’s Collaborative Combat Aircraft (CCA) program, which plans to deploy uncrewed jets alongside F-35s. Shield AI’s Hivemind autonomy stack is one of several candidates to power the CCA, alongside autonomy stacks from Anduril, Lockheed Martin, and others.
Can civilians use V-BAT?
No. V-BAT is a defense and government product. Civilian commercial drone work uses different platforms entirely — see our DJI vs Skydio vs Autel comparison for the civilian market.
The bottom line
Shield AI is the autonomous-flight software specialist of the new defense-tech wave. Founded by a former Navy SEAL solving a problem he saw in combat, the company built Hivemind into a credible candidate to power next-generation collaborative combat aircraft for the US Air Force and allied air forces. The $12.7 billion Series G valuation reflects expectations about Hivemind’s long-term position rather than just current V-BAT revenue.
The next 24 months are about whether Hivemind is in fact selected for the major CCA programs, whether V-BAT production scales to meet international demand, and whether Shield AI executes the developer-platform strategy through Hivemind Enterprise. All three of those will be tested.
For broader context: Anduril Industries Explained, AI in Military Drones: The Complete 2026 Overview, The Replicator Initiative Explained, White Stork: Eric Schmidt’s Stealth Drone Startup. Daily AI fundamentals in our free Beginners in AI newsletter.
Learn Our Proven AI Frameworks
Beginners in AI created 6 branded frameworks to help you master AI: STACK for prompting, BUILD for business, ADAPT for learning, THINK for decisions, CRAFT for content, and CRON for automation.
Get Smarter About AI Every Morning
Free daily newsletter — one story, one tool, one tip. Plain English, no jargon.
Free forever. Unsubscribe anytime.
Sources
- Shield AI official press release, Shield AI to acquire Aechelon, raise $2B at $12.7B valuation (March 26, 2026) — the primary-source company announcement of the Series G round, Advent International / JPMorgan Chase leadership, and the Aechelon acquisition.
- Fortune, Shield AI projecting more than $540M in revenue this year as valuation more than doubles to $12.7 billion (March 26, 2026) — independent financial-press coverage of the same round.
- Shield AI press release, Shield AI raises $240M at $5.3B valuation to scale Hivemind Enterprise.
- Wikipedia, Shield AI — background reference; current figures should be verified against Shield AI’s own press releases.
- The Next Web, Shield AI raises $2B at $12.7B for autonomous combat pilot Hivemind.
- U.S. Coast Guard official news at uscg.mil/News — the V-BAT maritime ISR IDIQ procurement (cited at $198 million) is announced and updated through official Coast Guard channels.
- Netherlands Ministry of Defence, defensie.nl English portal — official announcements of V-BAT procurement for the Navy and Marine Corps.
- DARPA Air Combat Evolution program, darpa.mil/program/air-combat-evolution — the official DARPA program page covering autonomous-combat-aircraft research, including the F-16 dogfighting experiments.
- U.S. Department of Defense Directive 3000.09, “Autonomy in Weapon Systems” (revised 2023) — available via the DoD Issuances portal at esd.whs.mil/DD. The governing policy framework for rules of engagement on autonomous systems including drone-fighter teaming.