On November 5, 2026, Shield AI raised $2.5 billion in a single financing package and announced it would spend up to $5.3 billion buying a radar company β pushing its valuation to $28.8 billion, more than double where it stood eight months earlier.
Shield AI makes the AI-pilot software the Pentagon is testing on everything from a runway-independent surveillance drone to a next-generation fighter jet. It is routinely described as the second-most-valuable company in defense tech, behind only Anduril. This deep dive covers what Shield AI actually builds, how it makes money, its full funding history, the Echodyne deal, leadership under new CEO Gary Steele, the V-BAT safety controversy, the competitive field, and what is and isn't known about an eventual IPO.

Company Snapshot
Figures from Shield AI's own funding announcements (via SiliconANGLE, Fortune, Washington Technology) as of November 2026. Shield AI is private and does not publish audited financials.
Founders
Brandon Tseng, Ryan Tseng, Andrew Reiter
HQ
San Diego, California
CEO
Gary Steele (since May 2025); former Splunk CEO
Headcount
Not disclosed; third-party trackers estimate roughly 1,500-1,900 in 2026
Key investors
Blackstone, Advent International, JPMorgan Chase, L3Harris, Hanwha Aerospace, a16z
Pending acquisition
Echodyne (radar), up to $5.3B, expected to close early 2027
What Shield AI Actually Does
Shield AI was founded in San Diego in 2015 by Brandon Tseng, a former Navy SEAL, his brother Ryan Tseng, and Andrew Reiter. The core product is not a drone β it is Hivemind, the AI-pilot software that flies inside the drone. The pitch is that modern combat environments jam GPS and sever communications links, so autonomy that depends on either fails exactly when it is needed most. Hivemind is built to keep flying and completing a mission without GPS and without a comms link back to a human operator.
That software currently flies on three of the company's own aircraft β the indoor-reconnaissance Nova 2 quadcopter, the runway-independent V-BAT vertical-takeoff surveillance drone, and the in-development X-BAT fighter aircraft β and has been adapted for the U.S. Air Force's Collaborative Combat Aircraft program, including flight-testing aboard Anduril's own YFQ-44A. The strategic bet is that Hivemind becomes a platform-agnostic autonomy layer licensed across many airframes, not tied only to Shield AI's own hardware.
How Shield AI Makes Money
Hivemind autonomy software
Licensing the AI-pilot stack to the Pentagon and to other aircraft makers, including a production award on the Air Force's Collaborative Combat Aircraft program and work powering Anduril's YFQ-44A. CEO Gary Steele has described this as one of three "fast-growing, high-margin" business lines the company is now organized around.
Aircraft hardware (Nova, V-BAT, X-BAT)
Selling and leasing its own autonomous aircraft directly to militaries. V-BAT sales were the company's largest disclosed 2025 revenue driver, per Forbes reporting, before the company said it fell short of its 2025 revenue target.
Government contracts and R&D
Contract revenue from U.S. and allied governments, including a $400 million combined public-private commitment to X-BAT (as of October 2026) and a reported $100 million in European contracts signed by mid-2026, plus pursuit of more than $1.8 billion in potential additional deals, per CFO Kingsley Afemikhe.
Mission Systems (simulation and, pending, radar)
A newer third line built from the Aechelon Technology simulation acquisition (closed June 2026) and, once the pending Echodyne deal closes in early 2027, in-house counter-drone radar β diversifying revenue beyond autonomy software and aircraft.
Funding History
Shield AI's valuation has compounded roughly 29x in five years β from just above $1 billion at its 2021 Series D to $28.8 billion today β with the steepest climb coming in the last 20 months, when the company more than quintupled in value across three rounds.
| Round | Date | Amount | Lead Investor(s) | Valuation |
|---|---|---|---|---|
| Series D | Aug 2021 | $225M | Not fully disclosed | ~$1B+ |
| Series E | JunβDec 2022 | $225M | Snowpoint Ventures | $2.3B |
| Series F | OctβDec 2023 | $500M | Hercules Capital (debt) | $2.8B |
| Series F-1 | Mar 6, 2025 | $240M | L3Harris, Hanwha Aerospace | $5.3B |
| Series G | Mar 26, 2026 | $2.0B | Advent International, JPMorgan | $12.7B |
| Series H | Nov 5, 2026 | $2.5B | Blackstone, JPMorgan Chase | $28.8B |
How we counted: each round's own announcement as reported by SiliconANGLE (Series F-1), Fortune (Series G), and Washington Technology (Series H); as of November 2026. The Series G figure includes $500M in non-dilutive preferred equity from Blackstone alongside $1.5B in equity; the Series H figure includes $1B in Series H equity, $1B in Blackstone structured equity, and $500M in JPMorgan-led debt.
Shield AI: Valuation Milestones, 2021-2026
Valuation climbed from roughly $1B at the 2021 Series D to $28.8B at the Series H five years later β a roughly 29x increase, with more than 5x of that coming in the last 20 months alone
SiliconANGLE, Fortune, and Washington Technology reporting on Shield AI's funding rounds, compiled November 2026
Data
| Item | Value |
|---|---|
| Series D ('21) | ~$1B+ |
| Series E ('22) | $2.3B |
| Series F ('23) | $2.8B |
| Series F-1 ('25) | $5.3B |
| Series G ('26) | $12.7B |
| Series H ('26) | $28.8B |
Product Portfolio
Hivemind
The core AI-pilot autonomy software stack: enables aircraft to fly and complete missions without GPS or a communications link back to a human operator.
Nova 2
A compact quadcopter for indoor reconnaissance and mapping in GPS-denied environments, deployed in hostage-rescue and urban operations since 2018.
V-BAT
A vertical-takeoff-and-landing surveillance drone that needs no runway or catapult β Shield AI's largest hardware revenue driver in 2025, and the subject of recent safety scrutiny (below).
X-BAT
An AI-piloted, low-observable, multirole VTOL fighter aircraft in development, backed by $400M in combined public and private commitments as of October 2026; flight testing is slated for later in 2026.
Aechelon Technology (acquired June 2026)
Geo-specific flight simulation and physics-based sensor training systems, used to train Hivemind in the Department of War's Joint Simulation Environment.
Echodyne (pending acquisition, ~early 2027)
Solid-state radar maker using patented Metamaterials Electronically Scanned Array (MESA) technology for counter-drone detection β will become an in-house sensing and counter-UAS business.
X-BAT: Betting the Next Decade on a Runway-Free Fighter
Shield AI's most ambitious bet is not a software deal β it is X-BAT, an AI-piloted, low-observable fighter aircraft designed to take off and land vertically from ships, islands, and austere sites with no fixed runway. The pitch is that long-range air combat power currently tied to a handful of vulnerable airbases and carriers could instead be launched from almost anywhere, flown by Hivemind rather than a human pilot sitting in the cockpit.
As of October 8, 2026, public and private commitments to X-BAT reached $400 million: the U.S. Navy put in $150 million through a modification to an existing Other Transaction agreement, which Shield AI matched with $150 million of its own capital, on top of an earlier $50 million Navy/Defense Innovation Unit commitment and a matching $50 million from Shield AI. About 500 engineers at Shield AI and suppliers including GE Aerospace are building flight-test aircraft, and the company says engine light-off, thrust-vectoring nozzle integration, and weapons-bay aerodynamics testing are complete, with flight testing slated for later in 2026.
X-BAT is the clearest answer to the question of what Shield AI is for beyond software licensing: if a runway-independent, AI-piloted fighter actually flies and the Pentagon buys it at scale, Shield AI stops being a subcontractor selling autonomy into other primes' airframes and becomes a prime contractor in its own right.
Revenue and Key Metrics
2024 revenue
~$267M
Sacra estimate
2025 revenue target
>$400M
missed, per Forbes
2026 revenue (projected)
$540M+
>80% YoY, company-stated
X-BAT commitments
$400M
as of Oct 2026
Shield AI does not publish audited financials. Research firm Sacra estimates 2024 revenue at roughly $267 million. The company had hoped to exceed $400 million in 2025 revenue and reach profitability, driven mostly by V-BAT sales, but CFO Kingsley Afemikhe confirmed to CFO Dive that the company missed its 2024 and 2025 financial targets and no longer expects 2025 profitability. Despite that, Tseng and Afemikhe told Fortune in March 2026 that Shield AI is projecting more than 80% revenue growth by the end of 2026 β implying revenue above $540 million for the year.
CFO Afemikhe said the company had signed roughly $100 million in European contracts and was pursuing more than $1.8 billion in additional potential deals as of mid-2026, telling CFO Dive, "While profitability timelines have adjusted, we are firmly back on track."
Competitive Landscape
Shield AI's closest and most-discussed rival is Anduril Industries, which competes directly for the Air Force's Collaborative Combat Aircraft autonomy work and is valued roughly 3x higher. General Atomics competes mainly as a legacy airframe maker β it builds the FQ-42A for the same CCA program and had a test-aircraft crash in April 2026 that it attributed to an autopilot error β while Skydio leads in small single-drone autonomy but has not scaled into CCA-class aircraft.
| Attribute | Shield AI | Anduril Industries | General Atomics |
|---|---|---|---|
| Founded | 2015 | 2017 | 1996 |
| Last priced/tender valuation | $28.8B (Nov 2026) | ~$90B (Oct 2026 tender) | Privately held; no disclosed valuation |
| Total disclosed funding (since 2021) | $5.7B+ | $11.3B+ (per site's own Anduril coverage) | N/A (not VC-backed) |
| 2025/2026 revenue | ~$267M (2024, Sacra); $540M+ projected 2026 | ~$2.2B (2025) | Not publicly disclosed |
| Core focus | Hivemind AI-pilot software + V-BAT/X-BAT aircraft + (pending) radar | Full-stack autonomous hardware + Lattice OS | Legacy unmanned aircraft (MQ-1/MQ-9, FQ-42A) |
| CCA program role | Mission autonomy provider (Hivemind); flying on Anduril's YFQ-44A | Prime contractor, builds YFQ-44A | Prime contractor, builds FQ-42A |
How we counted: each company's own funding or tender announcement as of November 2026. General Atomics is privately held and does not disclose financials; its row is included for CCA-program context only, not a valuation comparison.
Shield AI vs. Anduril: Valuation and Funding (2026)
Company funding and tender announcements, compiled November 2026
Anduril's ~$90B figure is from an October-November 2026 employee tender offer, not a primary funding round; a reported $100B round has not been confirmed.
Leadership Team
Gary Steele β Chief Executive Officer
Took over as CEO on May 13, 2025, after Shield AI announced the hire on March 12, 2025. Previously CEO of Splunk and president of go-to-market at Cisco. Steele has framed the company's growth around "three fast-growing, high-margin, mutually supporting business units" β autonomy software, aircraft, and now mission systems/sensing.
Brandon Tseng β Co-Founder, President & Chief Growth Officer
A former Navy SEAL and Surface Warfare Officer who served seven years in the Navy before co-founding Shield AI. Leads commercial growth, government relations, and the company's public advocacy for defense autonomy.
Ryan Tseng β Co-Founder, President & Chief Strategy Officer
Co-founded Shield AI and served as CEO from 2015 until the March 2025 leadership change, when he moved into a president and chief strategy officer role focused on senior government and industry relationships.
Kingsley Afemikhe β Chief Financial Officer
The company's public voice on financial targets and growth, including confirming to CFO Dive that Shield AI missed its 2024 and 2025 revenue targets while telling Fortune the company expects more than 80% revenue growth in 2026.
Bull Case / Bear Case
The Bull Case
- +Valuation has compounded roughly 29x in five years, with the sharpest jumps tied to specific rounds backed by strategic defense and financial investors (L3Harris, Hanwha, Blackstone, Advent, JPMorgan).
- +Hivemind is now flying on a rival's aircraft (Anduril's YFQ-44A) as well as Shield AI's own, suggesting the autonomy-software-as-platform thesis is working.
- +X-BAT has attracted $400M in combined government and company commitments before its first flight, and a working runway-independent fighter would make Shield AI a prime contractor, not just a software vendor.
- +The pending Echodyne acquisition diversifies revenue into counter-drone radar, a fast-growing category as drone threats proliferate globally.
- +CEO Gary Steele brings large-company scaling experience from Splunk, and the new financing brings in deep-pocketed, patient capital (Blackstone, JPMorgan) alongside existing venture backers.
The Bear Case
- -Shield AI missed its own 2024 and 2025 revenue targets (it had hoped to exceed $400M and reach profitability in 2025) and no longer expects to be profitable on the timeline it previously gave.
- -Reuters reported the V-BAT crashed more than 50 times over 18 months, and two separate incidents β a U.S. service member in April 2024 and a Romanian Navy officer in May 2026 β each partially severed fingers during operations.
- -A $28.8B valuation on a company still projecting roughly $540M in 2026 revenue (about 53x) is a rich multiple with no audited financials and a recent history of missed targets.
- -The Echodyne deal is contingent: if the acquisition falls through, the $2.5B financing round does not proceed as structured, and the deal still faces regulatory approval, including possible CFIUS review.
- -Shield AI is valued at roughly a third of Anduril's ~$90B tender mark and competes with it directly for the same Air Force CCA autonomy dollars.
Where We Could Be Wrong
The $5.7B+ total-funding figure is Value Add VC's own sum of disclosed rounds since 2021; it excludes an undisclosed seed-through-Series-C stretch from 2015-2021, so the company's true lifetime total raised is likely higher.
The 2024 revenue estimate (~$267M) comes from Sacra, a third-party research firm, not from Shield AI itself, and has not been independently verified.
Anduril's ~$90B comparison figure is an employee tender valuation, not a primary funding round, so it is not a perfectly like-for-like comparison with Shield AI's Series H.
IPO Outlook
Shield AI's IPO timeline is not publicly disclosed. As of November 10, 2026, no S-1 has been filed with the SEC, no underwriter or bank mandate has been reported, and neither Gary Steele nor Brandon Tseng has made an on-record comment about a target listing date. The closest thing to a market signal is the composition of the Series H round itself: Blackstone and JPMorgan Chase, both firms that routinely take pre-IPO positions in growth companies, just put a combined $2 billion into the business.
Six disclosed funding rounds in roughly five years, each larger and faster than the last, suggests Shield AI still finds private capital abundant enough that it hasn't needed to test public markets. It is consistently named alongside Anduril as a likely future defense-tech listing, but Anduril's own IPO timeline remains undisclosed too β meaning Shield AI has no real public comparable to benchmark against yet, let alone a stated plan of its own.
Bottom line: Shield AI turned a Navy SEAL's idea for an AI pilot into a $28.8 billion company in eleven years, with its valuation more than doubling twice in the last twenty months. The Echodyne acquisition and the X-BAT fighter program are real, expensive bets on becoming more than an autonomy-software vendor β but they come alongside a missed 2025 revenue target, an unresolved V-BAT safety record, and a valuation multiple that assumes the growth projections this time hold up better than last time's did. No IPO signal exists yet for Shield AI, just as none exists yet for Anduril, the company it is most often compared against.
All financial figures are based on publicly reported data, company announcements, and credible media reporting as of November 2026. Shield AI is a private company and does not publicly disclose audited financials or headcount; third-party revenue and headcount estimates cited here have not been independently confirmed by the company.
Latest from the Pulse
Get VC data most people never see
β free to subscribe
Trace's notes on venture, AI and startups, a few times a week. Join 5,000+ subscribers. No spam.
No spam, unsubscribe anytime. Sent via Substack Β· Privacy Β· or subscribe on Substack β