CX2

Over the last decade, the electromagnetic spectrum has become one of the most contested and least understood domains of contemporary warfare. The proliferation of low-cost unmanned aerial systems has fundamentally changed the battlespace, because these systems are inexpensive enough to deploy at scale, precise enough to threaten high-value assets, and expendable enough to overwhelm traditional air defense architectures.

Meanwhile, the radio frequency (RF) environment has become increasingly congested, and adversaries routinely jam or spoof GPS signals and disrupt communications. Most drones are spectrum-dependent machines before they are ever kinetic, relying on multiple communication links to their operators, which has made jamming those links one of the most effective counter-drone tools. As a result, the objective is shifting from destroying individual drones to identifying, geolocating, and neutralizing the operators and the RF networks that control them.

The economics of engagement help explain this shift. In the Red Sea, for example, the US routinely used interceptors costing in the low-to-mid millions of dollars per shot against drones costing orders of magnitude less. In Ukraine, low-cost drones have produced disproportionate battlefield effects, and as of March 2025, they caused an estimated 70-80% of Russian casualties. However, although the US spent roughly $5 billion on electronic warfare by May 2024, much of that capability remains embedded within high-cost platforms. Together, these trends make the ability to detect, classify, and precisely geolocate emitters in a contested electromagnetic environment a capability that could decide the next conflict.

CX2 builds airborne RF sensing payloads and drones that detect, classify, and geolocate the radio emissions connecting drones to their operators, along with jammers and other hostile emitters. Rather than competing in the market for kinetic counter-drone systems, the company treats drone warfare and electronic warfare as a single problem and focuses on finding the source.

CX2’s Vadris payload turns a commercial first-person view (FPV) drone into a seeker that points its pilot toward the operator controlling an enemy drone, and its Wraith drone maps hostile emitters across a battlefield. Both are designed to be cheap and simple enough for small tactical units rather than reserved for exquisite platforms. CX2's founders argued in October 2025 that control of the spectrum "will decide who sees, who acts, and who wins," and the company is building low-cost airborne sensors to give small units that control.

CX2 was founded in 2024 by Nathan Mintz (ex CEO), Mark Trefgarne (President), Porter Smith (CEO), and Lee Thompson (CTO). In 2023, Smith contacted Mintz after visiting Ukraine amid Russia's invasion of the country, where he had witnessed warfare shifting from a physical battlefield to an electromagnetic one. Instead of being fought primarily through traditional infantry, armored vehicles, and aviation units, the conflict increasingly revolved around drones, sensors, electronic attack, and spectrum control. US forces, meanwhile, relied on electronic warfare (EW) systems that Army Under Secretary Gabe Camarillo had said kept him up at night.

Smith had spent a decade flying AH-64 Apaches and MH-6 Little Birds with US Army Special Operations Aviation Command before becoming an Autonomy Team Fellow at the Defense Innovation Unit and then an investment partner at Andreessen Horowitz. He and Mintz later traveled to Ukraine together, and Mintz has said the lessons from that trip shaped CX2's work, along with his view that the defense industrial base had to abandon Cold War-era practices.

Smith approached Mintz because of his background in both EW engineering and defense company-building. Mintz spent 14 years as an EW and radar systems engineer at Raytheon and Boeing. In 2018, he co-founded Epirus, a software-defined directed-energy company focused on high-power microwave systems for counter-UAS applications; in 2020, he co-founded Spartan Radar, an automotive radar company.

Trefgarne complemented the defense-heavy technical bench with software company-building experience. He co-founded LiveRail, which CX2 describes as the world's largest video ad exchange, and sold it to Facebook (now Meta) in July 2014 for approximately $500 million. He then ran product for Facebook's AdTech group as a director of product management. Thompson brought RF hardware experience from SpaceX, where he led a team of RF engineers building the radio systems on Starship, giving the team hands-on experience with flight-critical RF hardware.

The four founders came together through 8VC's Build program, and CX2 became the program's fourth defense platform company. Mintz has described the ambition as building a 21st-century electronic warfare prime. Mintz led the company as CEO until July 2026, when Smith became CEO, and Mintz moved to a board and advisory role.

Product Overview

CX2 develops airborne RF sensing payloads, drones, and software that detect, classify, and geolocate wireless emitters across the battlefield. As of October 2026, its product line consisted of two named products. Vadris is an RF-seeking payload that mounts on an FPV drone and guides its pilot toward the operator controlling an enemy drone, and Wraith is an autonomous drone that maps hostile emitters such as jammers, drone control links, and ground stations across a wide area. Rather than relying on centralized EW platforms or expensive kinetic interceptors, CX2 distributes RF sensing onto low-cost tactical drones operating close to the edge.

Together, Wraith and Vadris form a complementary RF kill chain for tactical counter-UAS operations. Wraith serves as the upstream sensing layer, mapping the electromagnetic environment and geolocating hostile emitters. Once Wraith identifies an emitter, Vadris can be deployed downstream to home in on the specific control link and guide forces toward the operator behind the threat. The pairing supports a traditional Find-Fix-Finish workflow by detecting threats as soon as they transmit, geolocating both the drone and its operator, and handing off to downstream actions such as jamming, interdiction, or kinetic engagement. Wraith processes signals onboard using what CX2 calls its EW operating system.

Airborne RF Geolocation

CX2 was founded around an RF-first approach to counter-UAS. Rather than building another interceptor or jammer that defeats drones one at a time, the company seeks to identify the human operator controlling the threat by detecting, classifying, and geolocating the radio emissions that connect drones to their controllers. Neutralizing a single drone does not necessarily eliminate the broader threat, particularly when one operator can control multiple aircraft or quickly replace expendable platforms, and engaging each drone individually can be costly.

Compared with radar, cameras, and interceptors, this approach has three advantages. First, RF detection provides early warning at range and can cue other sensors toward a threat. Second, because the system passively monitors existing radio transmissions rather than emitting energy itself, it adds no signal of its own for an adversary to detect. Finally, by locating the pilot as well as the drone, it lets forces target the source of the threat rather than repeatedly engaging individual air vehicles.

Using RF emissions to locate adversaries is not a new concept. During the Cold War, anti-radiation missiles were developed specifically to locate radar transmissions. Systems such as the AGM-45 Shrike, first used during the Vietnam War in 1965, and later the AGM-78 Standard ARM and AGM-88 HARM, followed RF energy back to its source.

In 1973, DARPA's AXILLARY program explored cheap, small drones to attack enemy air defenses, and later systems like the IAI Harpy and AGM-136 Tacit Rainbow introduced the ability to remain airborne and search for emitters over time. These systems show that the physics of RF targeting has long been understood. What has changed is the enabling technology: software-defined radio architectures, smaller onboard computers, and the ability to deploy lower-cost sensing on small unmanned platforms.

Passive RF sensing is particularly well-suited to finding operators. Rather than transmitting signals like radar, passive RF systems operate purely as receivers, monitoring the electromagnetic spectrum to detect and analyze existing radio transmissions. By measuring characteristics of these signals, such as frequency, power, timing, and direction of arrival, passive sensors can infer the location and behavior of transmitters without emitting energy themselves.

Most existing passive RF detection systems, such as those from AeroDefense, Validrone, and Hidden Level, are ground-based, which imposes significant geometric constraints. Line of sight, the unobstructed path between transmitter and receiver, limits ground sensors to short detection ranges. For a sensor near the ground, the horizon sits only about three miles away, and terrain blocks much of what lies inside it. As a result, ground systems frequently struggle to observe operators positioned beyond the horizon or concealed by terrain.

Accurate RF geolocation typically requires measurements from multiple spatially separated sensors. A single stationary sensor can estimate the direction from which a signal arrives, but the transmitter could lie anywhere along that direction line, making it impossible to determine a unique location. To resolve this ambiguity, traditional systems rely on multiple sensors placed at different positions, comparing the direction or timing of the received signal to triangulate the emitter's position. However, deploying and coordinating distributed sensor networks increases system complexity, requires precise synchronization between sensors, and introduces additional sources of error from measurement noise and imperfect geometry.

Source: Inpixon

Placing the sensor on an airborne platform changes this geometry. A sensor operating at 100 feet above ground can observe transmitters within a radius of 12 miles or more, while a sensor one mile above ground has a radio horizon of 89 nautical miles, using the standard radio-horizon formula. That vantage point lets airborne sensors detect emitters, and potentially the operators behind them, that would otherwise remain hidden behind terrain or beyond the horizon of ground-based systems.

Motion further improves localization. As the drone moves, it collects measurements from multiple positions, each producing a line-of-bearing estimate toward the transmitter. Combining these measurements over time allows the system to triangulate the emitter's location, so the drone's motion substitutes for multiple stationary sensors and enables geolocation from a single airborne platform. By combining passive RF sensing with airborne mobility and onboard processing, this architecture extends detection range and enables earlier identification of drone operators rather than the drones themselves.

Source: Nathan Mintz’s Substack

Vadris

Source: CX2

Unveiled in November 2025, Vadris is CX2's RF seeker payload, designed to locate the human operator behind an FPV drone attack. The 1.2-pound payload mounts on commercial FPV drones, and CX2 says it requires about five minutes of operator training.

Vadris uses four antennas tuned to the frequencies and waveforms used by drones and ground operators to detect and characterize the control link between an operator and their drone in real time. Once it acquires a control signal, the system computes the direction from which the transmission is originating and overlays a live directional indicator directly onto the FPV pilot's video feed. As the drone maneuvers, Vadris continuously updates its line of bearing toward the signal source, helping operators narrow down the pilot's likely location on the ground and guide units toward confirmation and response.

CX2 reports that Vadris costs less than $10K and has detected control links at ranges of several kilometers. The design rests on the principle that if an adversary can control a drone at a given distance, Vadris can detect that control link as well. In October 2026, CX2 launched Vadris SDR, a software-defined version that detects the ExpressLRS, TBS Crossfire, and DJI OcuSync control protocols and can add others through software updates as adversaries change their RF signatures.

Wraith

Source: CX2

Unveiled in December 2025, Wraith is an autonomous electronic intelligence (ELINT) platform designed to detect and geolocate hostile emitters in environments where GPS, communications, and traditional unmanned systems often fail. Built for contested and electronically degraded battlefields, Wraith gives ground forces a clear picture of the RF spectrum, identifying GPS jammers, FPV jammers, drone control links, and ground stations, and turns that invisible electromagnetic activity into targeting data.

Wraith is a 32-pound Group 2 drone with about 45 minutes of flight time and a wideband RF front end, which lets it fly over the battlefield and detect hostile RF emissions. It is designed to operate autonomously in contested environments, using CX2's Pathfinder navigation software and GNSS hardening rather than relying on GPS.

Once airborne, Wraith geolocates emitters and produces a spectral heat map showing where signals originate, while an onboard electro-optical and infrared camera provides visual confirmation of targets. Its data can feed command-and-control platforms such as Anduril*'s Lattice and TAK, enabling a workflow in which units launch Wraith, map the spectrum, confirm the signal, and decide what to target, with Vadris available to home in on a specific control link. In testing, CX2 reports that Wraith maintained accurate geolocation in the presence of a 300-watt GPS jammer and delivered reliable detection beyond line of sight.

Customer

CX2's primary customer is the US military, specifically frontline tactical units operating in contested electromagnetic environments, where real-time spectrum awareness directly determines mission success. Rather than chasing the large, platform-level acquisition programs that traditional primes compete for, the company targets field-unit adoption with low-cost EW systems that units can field quickly. In May 2025, Mintz, then CEO, said that "the fight in Ukraine has shown that warfare has reached consumer scale." CX2's systems have been used or evaluated by units including the 4th Infantry Division's 4-10 Cavalry Squadron, the 75th Ranger Regiment through the Defense Innovation Unit's (DIU) Project G.I., the XVIII Airborne Corps during the Scarlet Dragon exercise, and a US Indo-Pacific Command (INDOPACOM) unit under a DIU contract awarded in April 2026.

CX2 has signaled ambitions to sell to allied militaries, with the US Department of Defense (DoD) remaining its anchor customer. Its selection in December 2025 for the Defence Innovation Accelerator for the North Atlantic (DIANA), NATO's accelerator, gives it access to defense ministries across the alliance, and Mintz said DIANA lets CX2 talk to "a lot of European MoDs at once." In July 2026, CX2 presented at a DIANA demo day in the UK. As of October 2026, the company had not announced a foreign sale. Beyond defense, CX2 markets Vadris for commercial and public-safety monitoring around stadiums, public events, critical infrastructure, and airports, and pitches Wraith for similar use around large public gatherings.

Although selling to the DoD has historically been challenging due to lengthy procurement cycles, recent acquisition reforms are more favorable to venture-backed companies like CX2. Army acquisition officials have acknowledged that "the pace of technology evolution is faster than the pace of traditional procurement processes," particularly in the electromagnetic spectrum. In April 2026, the Army established the Rapid Electromagnetic Warfare and Signals Intelligence (REWSI) initiative, which lets commanders field pre-vetted commercial EW and signals intelligence capabilities from a curated library rather than through traditional acquisition programs. The Army's FY2026 budget request also cited a need for more sophisticated countermeasures and deeper integration of AI into EW operations.

Market Size

CX2 competes within the EW market, specifically in airborne EW, counter-UAS, and RF sensing payloads for small unmanned systems. Estimates of the broader EW market vary, but they consistently point toward growth driven by rising geopolitical tensions, increasing defense spending, and the proliferation of unmanned systems. One report valued the global EW market at $21 billion in 2025 and projected it to grow from $22.1 billion in 2026 to $37.7 billion by 2036, a 5.5% CAGR over that period. Another estimated the market at $28 billion in 2025, growing to $64.7 billion by 2031 at a 14.9% CAGR, fueled by advances in AI and machine learning and increasingly contested electromagnetic environments. The US EW market was estimated at $8.4 billion in 2025 and projected to reach $14.2 billion by 2035.

The narrower counter-UAS market was estimated at $9.2 billion in 2026 and projected to reach $29.7 billion by 2031, a 26.5% CAGR from 2026 to 2031. Although much of this funding will continue flowing to large defense primes, CX2 addresses the part of the market focused on lightweight, attritable EW capabilities for tactical unmanned systems. Advances in miniaturized RF hardware and onboard computing have made it practical to put sensing payloads on small drones, where altitude extends the range of both monitoring and geolocation. The Army's FY2026 budget request allocated $1.7 billion in flexible funding across UAS, counter-UAS, and EW programs, including $79 million to buy ready-to-field EW capability.

CX2's competitors fall into two groups. RF detection specialists build passive detection and geolocation systems, mostly as fixed ground sensors, though Aaronia and Colibrex both sold drone-mounted payloads as of 2026. Large defense contractors and newer primes sell EW and counter-UAS systems that locate operators as one layer of a broader architecture, and several have moved toward small drones: Thales launched a passive RF payload for small drones in June 2025, and Anduril's Pulsar EW systems include an airborne configuration. CX2's distinction in both groups is that it puts RF sensing on small, inexpensive drones operated by tactical units.

RF Detection Specialists

Hidden Level: Founded in 2018, Hidden Level develops passive radar, drone detection, and RF sensing systems for defense, critical infrastructure, and public safety customers. The company had raised $130.9 million as of October 2026, including a $65 million Series C led by DFJ Growth in February 2025 at a $450 million post-money valuation, following a $35 million Series B, with investors including Lockheed Martin and Booz Allen Ventures. Like CX2, Hidden Level focuses on passive RF detection and emitter geolocation, but it deploys networks of fixed and vehicle-mounted ground sensors and sells them through a hardware plus data-as-a-service model, whereas CX2 delivers these capabilities through drone-mounted payloads operating at the edge.

Distributed Spectrum: Founded in 2020, Distributed Spectrum builds networks of low-cost RF sensors that use machine learning to classify and geolocate signals for defense customers. It had raised $25.2 million as of October 2026, including a $25 million Series A in March 2025 led by Conviction and Shield Capital, at an undisclosed valuation. Its sensors are distributed on the ground rather than mounted on drones, which makes it a closer analog to Hidden Level than to Wraith, though it shares CX2's focus on low-cost, attritable sensors for defense users.

Aaronia: Founded in 2003, Aaronia is a German manufacturer of RF spectrum analyzers, antennas, and drone detection systems. In June 2026, the company introduced the AARTOS Hawk T1, a passive airborne RF geolocation payload based on Blind Time Difference of Arrival (Blind-TDoA) technology that detects and locates emitters from a drone in real time. Aaronia is privately held with no recorded outside funding and no publicly disclosed valuation, and more than 600 AARTOS systems were in use worldwide by the end of 2024.

The Hawk T1 is one of the closest publicly known product analogs to Wraith, as both are lightweight airborne passive RF sensing payloads designed to detect and geolocate emitters from unmanned aircraft. However, Aaronia is a two-decade-old RF instrumentation company with an established equipment business rather than a venture-funded startup, and as a European supplier, it may face greater barriers to winning sensitive US defense programs where domestic manufacturing and trusted supply chains are prioritized.

Colibrex: Founded in July 2011 and now a wholly owned subsidiary of LS telcom, Colibrex specializes in drone-based RF measurement and spectrum monitoring. Its products include the NavAidDrone for inspecting and calibrating airport navigation aids and the LS OBSERVER Airborne Monitoring Unit (AMU), which provides drone-mounted spectrum monitoring, signal detection, and direction-finding. LS telcom, its publicly traded parent, had a market cap of €20.5 million as of October 2026.

The LS OBSERVER AMU is one of the closest product analogs to Wraith, as both use drone-mounted RF sensing to detect and geolocate wireless emitters from the air. However, Colibrex primarily serves civil aviation authorities and telecommunications regulators, such as Germany's air navigation service provider DFS and Brazil's telecommunications regulator ANATEL, whereas CX2 is built for tactical military operations in contested battlefield environments.

Incumbents and Primes

L3Harris: Formed in June 2019 through the merger of L3 Technologies and Harris Corporation, L3Harris Technologies provides EW, communications, sensors, and mission systems across every military domain. As of October 2026, the company had a market cap of $43.5 billion. Within its EW portfolio, Deceptor is a compact, software-defined EW payload for unmanned aircraft that detects and geolocates RF emitters and operates within L3Harris's DiSCO architecture for connecting EW systems, making it the closest analog to Wraith.

On the counter-UAS side, Wraith Shield competes most directly with Vadris, turning existing tactical radios into software-defined counter-drone systems that detect, classify, and jam FPV drones without additional hardware, an upgrade L3Harris said could eventually reach more than 100K fielded radios. L3Harris offers a far broader, fully integrated EW ecosystem backed by decades of defense relationships and large procurement programs, whereas CX2 focuses on lightweight, purpose-built airborne RF sensing payloads for tactical drones.

CACI: Founded in 1962, CACI International provides command-and-control, EW, and counter-unmanned systems to US government customers. Its counter-UAS portfolio includes BEAM 3.0, a backpack-portable electronic attack and direction-finding system that locates unmanned systems and their communications devices, and CORIAN, a fixed-site platform that fuses RF, radar, and electro-optical and infrared sensors to detect, identify, and track drones. CACI had a market cap of $12.9 billion as of October 2026.

From a mission standpoint, CACI is one of CX2's closest competitors, as both Vadris and CACI's systems prioritize locating the drone operator rather than only detecting the aircraft. The primary distinction is deployment: CACI delivers these capabilities through backpack-portable and fixed-site systems, whereas Vadris packages passive RF direction-finding into a payload mounted on tactical drones. CACI also benefits from longstanding government relationships, and in June 2026 the Pentagon's Joint Interagency Task Force 401 (JIATF-401) validated its SkyValor system for use across the Joint Force after border testing.

Thales: Founded in 1893, Thales is a French aerospace, defense, and security group that supplies radar, communications, and EW systems to militaries worldwide. It had a market cap of $49.8 billion as of October 2026. At the 2025 Paris Air Show, Thales introduced a compact EW payload for small drones that weighs under 11 pounds, draws less than 40 watts, and passively scans for, detects, and locates enemy radio emitters, either free-flying or tethered. That makes it one of the closest product analogs to Wraith among the large primes. Thales has decades of procurement relationships across NATO militaries, the market CX2 is trying to reach through DIANA. As a European supplier, though, it faces the same barriers to sensitive US programs as Aaronia, where CX2 builds its systems in El Segundo.

Neoprimes

Epirus: Founded in 2018, Epirus builds Leonidas, a solid-state high-power microwave system that Epirus says has proven effective in Department of Defense testing against drones, drone swarms, and other electronics. The company had raised $537.6 million as of October 2026, the latest a $250 million Series D in March 2025 led by 8VC and Washington Harbour Partners at an undisclosed valuation. Mintz co-founded Epirus before starting CX2, and 8VC, which co-led CX2's seed round, also led Epirus's Series D. The two companies sit at opposite ends of the counter-UAS kill chain: Leonidas disables drones in flight, while CX2's products find the drones and the operators controlling them. That makes Epirus less a like-for-like competitor than a rival for the same counter-drone procurement budgets, and potentially a complementary layer in the same defensive architecture.

Anduril: Founded in June 2017, Anduril builds autonomous defense systems for air, land, sea, and space, tied together by its Lattice command-and-control software. The company had raised $11.4 billion as of October 2026, the latest a $5 billion Series H in May 2026 led by Andreessen Horowitz and Thrive Capital at a $61 billion post-money valuation. Its Pulsar line of EW systems comes in vehicle, fixed-site, and airborne versions, and in April 2025 Anduril introduced Pulsar-L, a shoebox-sized system under 25 pounds, available in airborne and expeditionary configurations. In an autonomous mode, it reads the spectrum, decides which signals are threats, and engages them. Pulsar and CX2's products both sell AI-driven RF capability to tactical units, but Pulsar is built mainly to detect and jam, while CX2 focuses on passively geolocating emitters and their operators. Anduril also controls the command-and-control layer CX2's products plug into: Wraith is compatible with Lattice, which makes Anduril both a competitor and a platform CX2 depends on.

CX2 generates revenue through hardware and software sales to government customers, primarily via DoD procurement contracts, DIU awards, and partnerships with other defense companies and systems integrators. Public pricing remains limited. Vadris is reported to cost less than $10K, works across drone platforms, and requires about five minutes of operator training, while pricing for Wraith has not been disclosed. CX2's $2 million DIU award in April 2026 covered 30 Wraith systems and 25 units of an undisclosed RF deception product, an average of $36.4K per unit across the two products.

As a hardware-plus-software company, CX2 is more asset-intensive than a pure software startup, although it remains considerably lighter than traditional defense primes. The company manufactures Wraith at its El Segundo facility using US and allied components, which requires investment in manufacturing capacity and supply chains that many early-stage defense startups delay. It also invests in R&D across multiple product lines, including Vadris, Wraith, and the undisclosed RF deception product, while bearing the bill-of-materials costs of RF sensors, drones, and embedded electronics. As a result, gross margins, particularly on lower-cost products such as Vadris, are likely to depend on manufacturing scale, component sourcing, and production volume.

Government Contracts

CX2's traction has come through DoD contracts and military development programs. In April 2026, DIU selected CX2 for Project G.I. Design Reference Mission 3, aimed at adoption by an INDOPACOM unit, and awarded it a $2 million procurement contract for 30 Wraith systems and 25 units of an undisclosed RF deception product. Before this award, DIU had downselected Vadris in November 2025 for Project G.I. Design Reference Mission 2 ("Accelerate the Kill Chain"), paired with Neros's Archer drone, while Wraith was developed through a Cooperative Research and Development Agreement with a US government component and flown in multiple military evaluations. By October 2026, CX2 reported that more than 450 Vadris systems had been fielded to the US military.

Operational Demonstrations & Partnerships

Beyond direct contracts, CX2 has tested its technology in military exercises and integrated it into broader defense ecosystems. During the Army's Exercise Ivy Mass at Fort Carson in May 2026, soldiers from the 4th Infantry Division's 4-10 Cavalry Squadron operated Wraith without contractor support, flying nearly a dozen autonomous missions over two weeks to geolocate RF threats as part of the division's Next Generation Command and Control initiative.

In March 2026, CX2 partnered with Neros to integrate Vadris with Neros's Archer drone, creating a modular counter-UAS system and aligning the two companies' product roadmaps and go-to-market plans. In December 2025, CX2 joined the Picogrid Partnership Ecosystem, which connected its passive RF sensors into Army command-and-control systems during the Scarlet Dragon exercise. Internationally, by March 2025, CX2 had partnered with Ukrainian manufacturers including Skyfall, whose bomber drones carry CX2 software and sensors, giving it feedback from one of the world's most active EW theaters.

As of October 2026, CX2 had raised a total of $46 million, and its valuation had not been publicly disclosed. In May 2025, the company raised a $31 million Series A led by Point72 Ventures, with participation from Andreessen Horowitz, 8VC, Pax Ventures, 201 Ventures, and Upfront Ventures. That followed a $15 million seed round in 2024 led by Andreessen Horowitz and 8VC. The Series A was roughly twice the size of the seed and closed about a year after the company's founding. As a member of NATO DIANA's 2026 cohort, CX2 is also eligible for the program's €100K first-phase non-dilutive grant.

Growing Counter-UAS Budgets

Counter-drone capabilities have become a central Pentagon investment priority. The FY2027 defense budget requests more than $70 billion for military drones and counter-drone systems, which the DoD comptroller in April 2026 called the largest investment in these technologies in US history, after officials sought $13.4 billion for autonomous systems and $3.1 billion for counter-UAS in FY2026. Within that, the Army has proposed nearly $1 billion for small counter-UAS procurement in FY2027, including $108 million for handheld, wearable, and dismounted systems, a category aimed at the same dismounted units that operate Vadris.

Counter-UAS as Permanent Infrastructure

Counter-UAS is evolving from an urgent wartime capability into a permanent layer of military infrastructure, which increases the long-term value of platforms that become embedded in it. In March 2026, the Army agreed to an enterprise contract with Anduril worth up to $20 billion over 10 years, and JIATF-401, the DoD's central authority for counter-UAS, selected Anduril's Lattice as its enterprise counter-UAS command-and-control platform. Both reflect a shift toward long-term modernization instead of one-time equipment purchases. Wraith is already compatible with Lattice, so it can feed the command-and-control layer the Army and JIATF-401 have standardized on, which is a prerequisite for being bought at enterprise scale.

Faster Acquisition Pathways for Startups

Defense acquisition reform is making it easier for venture-backed startups to win DoD contracts. Through organizations such as DIU, the DoD increasingly relies on rapid acquisition pathways designed for nontraditional defense companies, and DIU can award prototype agreements in as few as 60 to 90 days. Obligations under Other Transaction Authority agreements for prototyping and production grew from $1.8 billion in FY2016 to more than $18 billion in FY2024. Initiatives such as DIU's Blue UAS List, a catalog of vetted drones and components managed by the Defense Contract Management Agency since December 2025, aim to make it easier for qualified vendors to sell to the military, and CX2 markets Vadris as Blue UAS compliant. CX2 has already won two Project G.I. selections through DIU, which gives it a track record to build on as these pathways scale toward larger production awards.

RF-Silent Drones

CX2's products detect and geolocate drones through their RF emissions, but drone technology is reducing reliance on continuous RF communications. Fiber-optic drones replace the radio control link with a physical cable, removing the signal that RF detection and jamming systems exploit, and the Army has said the signals these drones do emit are "extremely weak" and difficult to detect. At the same time, battlefield experience in Ukraine has driven adoption of AI-enabled guidance that lets drones keep targeting even when the operator's link is jammed or lost. If adversaries continue shifting toward RF-silent, autonomous, or fiber-tethered platforms, CX2's technology could become less effective against the most sophisticated threats while remaining useful against RF-controlled FPV drones, which shortens the window for the company to expand beyond pure RF sensing.

Dependence on Federal Budget Timing

CX2's revenue is concentrated among defense and government customers, making growth dependent on federal procurement cycles rather than commercial demand. Defense contracts move on Congress's budget calendar, and when appropriations are delayed, agencies operate under continuing resolutions (CRs) that generally hold spending at prior-year levels and restrict new program starts. The DoD has operated under a CR in 37 of the last 49 fiscal years, and FY2025 was the first year since at least FY1977 in which the DoD operated under a full-year CR, forcing the department through an entire fiscal year without a new budget.

These delays can postpone contract awards, slow procurement and R&D spending, and increase program costs. Because CRs restrict new starts, they weigh most heavily on programs that did not exist in the prior year's budget, which is where many startup contracts sit. While long-term defense spending remains supportive, the timing and predictability of CX2's revenue may remain volatile as long as the company relies primarily on government customers.

Manufacturing at Production Scale

While rapid-acquisition programs have lowered barriers for startups to win early prototype contracts, long-term success increasingly depends on the ability to manufacture systems reliably, quickly, and at scale. The DoD is shifting toward commercial procurement models that prioritize affordability, production speed, and the ability to produce low-cost systems in volume.

This shift has fueled the emergence of a generation of new primes such as Anduril and Helsing. Those companies, along with peers such as Saronic, differentiate themselves through software and autonomous systems and through manufacturing strategies designed to deliver systems quickly and at scale. In August 2024, Anduril announced Arsenal-1, a hyperscale manufacturing facility later sited in Ohio and designed around simplified, modular product design and commercial supply chains. In July 2026, Saronic announced $3 billion in planned investment in Port Alpha, a shipyard in Brownsville, Texas, for building autonomous vessels.

New acquisition programs increasingly reflect these priorities. As planned in February 2026, the DoD's Drone Dominance program narrows its vendor pool over successive phases, from 25 Phase I competitors, about 12 of which receive orders, to three to five production suppliers, as order volumes rise from 30K to 150K drones and the target unit cost falls to about $2.3K. At that stage, competitive advantage shifts from demonstrating technical performance to proving production capacity and supply chain resilience.

Although CX2 has demonstrated technical capability and early customer traction, including more than 450 Vadris systems fielded as of October 2026, it has yet to prove it can manufacture and deliver systems at the scale major production programs require. In May 2026, the company opened a 30K-square-foot headquarters and manufacturing center in El Segundo. Making that transition will require substantial investment in manufacturing infrastructure, supplier relationships, and operational execution, a transition where many early-stage defense startups struggle.

CX2 builds airborne RF sensing payloads, drones, and software that let small military units detect, classify, and geolocate radio emitters, including the operators controlling enemy drones. Its two products, the Vadris seeker payload and the Wraith survey drone, are designed to be cheap and simple enough for tactical units rather than reserved for exquisite platforms. The US military is its primary customer, and NATO's DIANA program gives it a route to allied defense ministries.

As of October 2026, CX2 had fielded more than 450 Vadris systems, won two DIU Project G.I. selections, and raised $46 million. The open questions are whether it can scale production fast enough to compete for large procurement programs against primes such as L3Harris and CACI that sell into the same mission, and whether its RF-first approach keeps pace as adversaries shift toward fiber-optic and autonomous drones that emit little or no radio signal.

*Contrary is an investor in Anduril through one or more affiliates.