The NatSec100 and the Patent Paradox

In the 2026 NatSec100, the companies winning on execution and the companies building the deepest patent positions are often not the same. The gap between them is where long-term value will be decided in the next wave of defense tech consolidation.

The 2026 NatSec100 changed what it measures. For the first time, Silicon Valley Defense Group folded U.S. government contracting data into its momentum scoring, moving the emphasis from innovation potential toward demonstrated execution: selling, fielding, and producing rather than merely building. The resulting list is a clear read of where traction sits in defense tech today.

However, patents speak to the next wave of value proposition and leverage. A ranking built on contracts, capital, and headcount reflects who is winning now. A company's patent portfolio defines its exclusive domain when competition shows up. Imitation is the highest form of flattery, and markets move to copy imitators.  

To compare a market vs. patent methodology, we attributed every U.S. patent family held by a 2026 NatSec100 company, granted or in examination, to the operating entity behind it, which included more than 1,700 patent families across the cohort. Within this framework, we identify disparities in the balance of power between market success and exclusive rights. These disparities identify opportunities and risks in the defense and intelligence sectors.  

The two lenses do not line up — but that is the useful part.

The shape of the cohort's IP

Slightly more than half of the 100 companies hold attributable patent families as of late May 2026. The remainder hold none that surface in the public record. Absence is not automatically weakness — early-stage, software-forward, and classified companies often rely on speed and secrecy rather than patent filings — but the lack of IP is measurable, and introduces risk. Of the families that exist, the breakdown between granted patents and pending applications splits almost evenly. The 2026 NatSec100 is a cohort that is still actively filing, not one resting on an issued core.

The depth is concentrated. The 10 largest portfolios hold roughly two-thirds of all families in the cohort, and they cluster in capital-intensive, disclosure-exposed fields — launch and space hardware, advanced materials, additive manufacturing, photonics, autonomy, and quantum — where a fielded product can be examined and protection is therefore worth paying for. The leaders span a range of strategic archetypes:

Archetype	Typical NatSec Rank Range	Patent Family Count Range	Pattern
Execution Leader / Balanced IP	Top 25	50–180	Strong contract performance supported by moderate to strong patent position in hardware-intensive domains
IP Fortress / Contract-Light	30–90	100–180	Deep patent estates in long-cycle sectors (materials, manufacturing, photonics) that have not yet converted to major contract wins
Software-Forward / Light IP	Top 15	10–90	High execution velocity in artificial intelligence (AI), data platforms, or cyber with minimal formal patent protection; often relies on trade secrets
Emerging Filer / Momentum Builder	20–40	30–70	Rapidly expanding pending applications, signaling preparation for financing, partnerships, or exits
Zero-Patent Execution Leader	Top 10	0–15	Software, classified, or service-oriented firms achieving top-tier government traction without formal intellectual property (IP) filings

Archetypes derived from analysis of patent families held by 2026 NatSec100 companies, granted or in examination. A family is counted once and treated as granted if any member has issued. 

Execution and IP pull apart

Read down the patent column and the NatSec100 ranking varies. To illustrate, a leading launch provider, ranked in the middle quartile on execution, holds one of the largest portfolios in the cohort. Advanced manufacturing and photonics firms ranked outside the top 50 carry more families than all but a handful of higher-ranked peers. Meanwhile, the top of the execution list is uneven on IP: The top-ranked autonomy and defense systems integrator holds a real but mid-sized portfolio; a top-five space hardware company holds a modest collection; a top-five AI platform holds fewer than 15 families; a top-five maritime autonomy firm holds only pending applications; and another top-five autonomy leader holds none that appear in the public record.

Two patterns may explain most of this. The first pattern relates to those execution leaders with light portfolios. For a software-forward or classified company, a patent is a public teaching document. Declining to forego patent protection early on historically carried little harm. However, what works early becomes a liability at scale. The more widely a product is deployed, the more it can be reverse engineered, shared across partners and suppliers, and carried off by departing staff, leaving an unpatented advantage with nothing to defend it. Patent protection that seems “optional” at the prototype stage can make the difference between establishing dominance and ceding a competitive advantage once the market adopts a technology.

The second is patent-rich companies sitting lower on the execution list — firms in advanced manufacturing, photonics, robotics, AI automation, and networked space systems. These companies tend to operate on longer production or regulatory timelines, where IP accumulates ahead of contract visibility. For strategic acquirers or primes, this group offers latent optionality: control over technical pathways that have not yet shown up as revenue.

Read the pipeline, not just the portfolio

Across the whole cohort, annual filings climbed from roughly 110 in 2021 to more than 400 in 2024, reflecting annualized growth of roughly 54%, before the normal publication lag obscures the most recent year. But the shape of a portfolio matters as much as its size, and the cleanest “tell” is the ratio of pending applications to granted patents. Because an application precedes its grant by one to four years, a pending-heavy estate is a forward indicator — a company actively widening its protected ground ahead of an inflection point such as a new platform, a financing round, or an acquisition.

On that measure, the accelerating names stand out: A major additive launch provider files more than four pending applications for every granted patent; space hardware and lunar infrastructure firms show ratios of three to four to one; quantum computing and frontier AI companies show ratios better than two to one. At the other end sit granted-heavy or quiet pipelines — some top-tier AI platforms show double-digit granted patents with nothing pending; blockchain intelligence and autonomy simulation firms show similarly lopsided ratios; and even leading defense integrators and materials innovators show portfolios weighted toward the issued side. For a mature core, this profile is understandable. For a company that is still scaling and developing, a pipeline that has gone quiet posits the question: Is this a deliberate shift to trade secrecy or have innovations and filings simply lapsed?  

Readers should note that the patent system can be faster than the contracting cycle in some instances. For example, patent applications in the U.S. Patent and Trademark Office’s Track One program can now be allowed in three months and granted in four months. Historically, patent examination took 2-5 years.  

Where it gets decided: The deal table

Patent strength in this market is always priced and leveraged even if less common to litigate. When a few companies hold leading patent positions, they set the terms on which everyone else integrates. A firm with depth can cross-license through conflict or trade access for access with a prime; a firm without patents negotiates from weakness or pays a premium to participate. As the cohort moves toward consolidation, prime build-versus-buy decisions, and deeper production partnerships, companies with exclusive rights are better positioned to secure value for their innovations.

That is the practical content of the four questions every acquirer and lead investor now asks and that every founder should be able to answer: 

• Freedom to operate. Can the company ship its current and planned products without infringing someone else’s claims? In crowded technology fields such as autonomy, radio frequency, and space hardware, patent clearance must be performed and cannot merely be assumed.
• Owning the innovation edge. Do the claims protect what the company builds today and intends to build next, or do they protect an idea it has moved past? Scope is read against the product, not counted.
• Chain of title. Is the IP cleanly assigned to the operating entity, with inventor and contractor agreements captured? Gaps are common in fast-grown startups and expensive to fix later.
• Government rights. What march-in, data, and license rights has the customer acquired through Small Business Innovation Research, Other Transaction Authority, or procurement terms? In defense these can quietly reshape what an acquirer is actually buying.

What to do with this information

For founders

Match your filing strategy to your field and stage. If you build in a disclosure-exposed domain — e.g., launch, materials, photonics, additive manufacturing, RF — a thin or flat portfolio is a liability that compounds as you field product. File ahead of deployment, not after. If you are software-forward or doing classified work, decide deliberately between patent and trade secret and document the choice, because an acquirer will ask. Keep chain of title clean by assigning to the operating entity and capturing inventor and contractor IP from the start. Be able to produce a one-page map tying claims to products. And watch your own pipeline: A pending-heavy estate reads as momentum in diligence, while a pipeline that went quiet while you kept shipping reads as risk.

For investors  

Read the portfolio's shape, not only its size. A high pending-to-granted ratio at a capital-intensive company is a forward signal worth underwriting; zero pending at a company still scaling raises questions but is not necessarily a verdict. Look hard at the patent-rich names sitting lower on the execution list — advanced manufacturing platforms, photonics interconnects, robotics systems — as control points over pathways that have not yet converted to revenue. Budget diligence to attribute families to the operating entity and read the claims against the current product before forming valuations.

For acquirers and primes  

In consolidation, the firms holding the defensible positions in a domain set the integration terms. Map the choke points — who holds the families in counter-unmanned aircraft systems, photonic interconnect, and additive structures — and weigh IP control alongside revenue in any build-versus-buy decision. The cheapest time to secure a technical pathway is before a competitor decides it wants the same one.

For the ecosystem

Execution is what the Pentagon needs and what capital is finally rewarding. But execution sets up the contest; it does not settle it. The companies that will still hold their ground when these technologies are common — i.e., when the second and third entrants arrive and the primes decide what to build versus buy — are the ones building defensible, well-documented IP positions now.

A note on method

The figures here are U.S. patent families, granted or in examination, attributed to each 2026 NatSec100 company's operating entity from current-assignee records. Counts are at the family level rather than the individual publication: A family is counted once and treated as granted if any of its members has issued. Otherwise, it is treated as pending. Attribution to the operating company — resolving subsidiaries, holding entities, and assignment transfers — was done by hand; assignees belonging to unrelated companies that share a name were set aside. Slightly more than half of the 100 companies had attributable families. Absences may reflect early stage, a software or trade-secret strategy, classified work, or IP held under structures not captured in public assignee data, and are not necessarily measures of quality. The official NatSec100 ranking, produced by Silicon Valley Defense Group, uses a momentum methodology drawn from public data including federal contracting, capital formation, and headcount. Nothing here assesses the validity, scope, or value of any specific patent; establishing those requires reading the claims.