Sign in Start free trial

Sign in Start free trial

Shield AI

Shield AI is an Autonomous Aviation company founded in 2015 and based in San Diego, United States. It has raised $3.6B in total funding, most recently a Series G in 2026 at a $12.7B valuation.

Shield AI develops Hivemind, an AI pilot platform that enables aircraft and drones to operate autonomously in GPS- and communications-denied environments. Its flagship products include the V-BAT vertical-takeoff unmanned aircraft and Nova indoor reconnaissance quadcopter, both fielded by the U.S. Department of Defense and allied militaries. Founded in 2015 in San Diego by former Navy SEAL Brandon Tseng, his brother Ryan Tseng, and Andrew Reiter, Shield AI has positioned itself as the leading pure-play autonomy software provider for defense aviation. In 2025 it launched Hivemind Enterprise, opening its autonomy stack to OEMs and governments building their own unmanned systems. In March 2025 Shield AI raised $240M at a $5.3B valuation, with Cisco veteran Gary Steele joining as CEO and Ryan Tseng transitioning to President and Chief Strategy Officer. In March 2026 Shield AI raised $2B at a $12.7B valuation in Series G funding led by Advent International, alongside $500M in preferred equity from Blackstone, more than doubling its prior valuation. The company projects $540M+ in revenue for 2026. Total capital raised exceeds $3.6B.

Shield AI is building the autonomy layer that could become the de facto operating system for Western military drones, a strategic capability at the center of modern air power doctrine.

shield.ai Autonomous Aviation

United Statesprivate

At a glance

Total raised

$3.6B

last round series g

Headcount

1001-5000

approximate · public range

Headquarters

San Diego

United States

Founded

2015

11 years active

Status

Private

Autonomous Aviation

Latest valuation

Funding history

Total raised

$3.6B

Rounds tracked

1

Latest stage

Series G

Latest valuation

$12.7B

Date	Stage	Amount	Valuation	Lead investors
Mar 6, 2025	Series F	$240M	$5.3B	L3Harris, Hanwha Asset Management

How they make money

AI bull / bear

Bull case

Shield AI owns the critical autonomy choke point as military aviation pivots from crewed jets to autonomous swarms. Hivemind's GPS-denied operation is combat-proven with U.S. and allied forces—a barrier to entry few competitors can match. The V-BAT and Nova platforms already generate defense revenue while the March 2025 launch of Hivemind Enterprise turns proprietary software into a licensable layer for OEMs, dramatically expanding TAM without capital-intensive manufacturing. The company raised $1.4 billion through Series F, and recent intelligence indicates a $2 billion Series G in March 2026, signaling sustained investor conviction and runway to scale. Strategic investors L3Harris and Hanwha Asset Management bring procurement channels and international partnerships. New CEO Gary Steele—Cisco veteran—adds enterprise go-to-market muscle precisely as Shield pivots to platform licensing. With Western governments racing to field autonomous air power and China advancing its own systems, Shield's dual advantage in software and deployment creates a compounding moat.

Bear case

Shield AI faces a crowded autonomy race with deep-pocketed rivals. Anduril, valued higher and flush with capital, is building overlapping capabilities and has secured major DOD programs. Traditional primes (Lockheed, Boeing, Northrop) are embedding autonomy in-house, threatening to bypass third-party software layers. The Hivemind Enterprise pivot to licensing is unproven—OEMs may resist ceding control of core autonomy IP, preferring proprietary stacks or open-source alternatives. Defense procurement is glacial and lumpy; even with fielded systems, scaling revenue depends on multi-year budget cycles vulnerable to political and fiscal shocks. The CEO transition mid-scaling introduces execution risk, and the company's $5.6 billion valuation demands blockbuster outcomes that hinge on becoming the Windows of military drones—a winner-take-most scenario far from guaranteed. International expansion faces export controls and geopolitical friction. If Hivemind doesn't achieve platform lock-in soon, Shield risks becoming a niche hardware vendor squeezed by integrators.

Generated by claude-sonnet-4-5-20250929 · April 25, 2026Refreshes quarterly · see disclosures

Products

Hivemind (AI pilot)
AI autonomy software that assumes the role of a human pilot, enabling unmanned defense systems to sense, decide, and act. Unlike traditional autopilots that simply follow preplanned routes, Hivemind can reroute around no-fly zones, avoid or engage obstacles, respond to unexpected conditions, and complete missions safely without human intervention. In continuous combat deployment since 2018, making it the most operationally experienced autonomous AI pilot. Has piloted 26 classes of vehicles including F-16s, jet-powered UAVs, helicopters, drone boats, and ground vehicles. Recently demonstrated autonomously flying two Navy BQM-177A aircraft during a December 2025 flight test at Point Mugu Sea Range in California.
Hivemind Enterprise
Autonomy software suite for developers and organizations building, testing, evaluating, and deploying autonomy for intelligent machines. Accelerates autonomy development by delivering a multi-year head start with platform products, AI-powered toolsets, and leveraging Shield AI's proven edge autonomy work enabling unmanned systems to operate in GPS- and comms-denied environments. The product suite includes an integrated autonomy factory, industrial-grade production ready middleware, a rich autonomy catalogue, and seamless mission control. The same tools that make up Hivemind Enterprise have already enabled Shield AI to autonomously fly F-16s, MQ-20s, MQM-178s, and operate V-BATs in GPS- and communications-denied environments.
V-BAT
Vertical take-off and landing (VTOL) reconnaissance unmanned aerial vehicle with ducted-fan design, more than 12 hours of endurance, and heavy-fuel engine. Designed, tested, and deployed for electronic warfare environments, V-BAT operates in GPS-denied and comms-contested conditions using Shield AI's Hivemind autonomy software, which includes visual odometry navigation capabilities. 40 lbs maximum payload capacity; lands in zones as small as 4.6 m × 4.6 m. First unmanned cargo delivery to a U.S. Navy ship at sea (December 2022).
Nova quadcopter
Compact quadcopter designed for indoor exploration, mapping, and reconnaissance in GNSS-denied environments, running Hivemind autonomy software to operate without GPS or human operator control. Deployed in high-stakes missions since 2018, including Israeli hostage rescue operations. Nova 2, the current generation, features an embedded Xavier AGX compute module and 15-camera sensor array, enabling dense depth matching and feature tracking for robust state estimation in complex 3D indoor spaces. The platform serves as the foundational testbed for Shield AI's broader Hivemind autonomy stack.