Lockheed Martin generates roughly $70 billion in annual revenue. It is the largest defense contractor on earth by every conventional measure, contract value, employee count, program portfolio, congressional footprint.
Type "best autonomy companies for U.S. military," or "leading defense AI contractors," or "top next-generation warfare technology vendors" into ChatGPT, Claude, Perplexity, Gemini, or Google AI Overviews. Lockheed Martin frequently does not appear in the top three answers. Often it does not appear in the top ten.
That is the citation gap. And it is the most consequential communications problem in the legacy defense industry. The pattern repeats across categories, see the automotive Citation Share study for the same dynamic playing out in autos.
The 28,400-prompt finding
The 5W Defense & Aerospace AI Visibility Index 2026 tested 28,400 prompts across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. The finding that matters most for defense communicators: the companies winning AI citation share are not the companies winning the contract dollar leaderboard. Anduril and Palantir appear in defense-discovery prompts at rates dramatically higher than their share of defense industry revenue, and Lockheed Martin ranks third in defense citation share, behind both.
Citation Share is the new market share. In defense, that's now literally true. Program managers, congressional staffers, allied-government procurement officers, defense-tech investors, and trade reporters now run preparatory research through generative AI before they engage primary sources. A company not in that answer doesn't get evaluated. A company in that answer with stale framing gets evaluated against stale framing.
Market leadership is not citation leadership
Recent AI visibility research has documented the pattern across those 28,400 prompts and five generative AI platforms. Legacy primes, Lockheed Martin in particular, appear less frequently in AI-generated responses than their market scale would predict. Defense-tech challengers, Anduril, Palantir, Shield AI, appear more frequently than their revenue would predict.
The gap is not capability. Lockheed builds the F-35, the most advanced operational fighter in the world. It produces the THAAD missile defense system, the PAC-3 interceptor, the Aegis combat system, and the Skunk Works lineage that produced the U-2, the SR-71, the F-117, the F-22, and a long list of programs that defined modern American airpower.
The gap is communications structure. The categories where Lockheed loses citation share, autonomy, drone swarms, AI software defense, next-generation warfare, defense-tech innovation, are exactly the categories where buyer-intent prompts now concentrate. And in those categories, Lockheed's content does not retrieve.
Where Lockheed wins (and why)
The research found that Lockheed still surfaces consistently in several prompt categories:
- F-35 program retrievals, multi-decade media coverage, primary-source dominance, no comparable challenger.
- Hypersonic and long-range fires, the LRHW (Dark Eagle) and related programs produce retrievable anchors.
- Missile defense, THAAD, PAC-3, Aegis. Decades of trade press.
- Space and satellites, the Lockheed Martin Space portfolio is well-documented.
These are not small categories. They are the legacy strengths. The communications problem is not that Lockheed has lost ground on its existing categories. It is that the categories defining the next decade, autonomy, AI software, swarm warfare, Collaborative Combat Aircraft, defense-tech challenger ecosystem, are not Lockheed-coded inside the retrieval models.
Where Lockheed under-cites (and why)
Lockheed has serious work in nearly every category where the data shows it under-citing. The CCA selection went to Anduril and General Atomics, but Lockheed has been deeply involved in autonomous systems for decades. Lockheed Martin Skunk Works has the SR-72 hypersonic concept, the Stalker UAS, the Indago drone family, and ongoing autonomous-ground-vehicle work. The capability exists.
The capability is not communicated in the structure the AI engines retrieve. Lockheed's autonomy work appears in dense corporate press, IR-friendly investor materials, and capability brochures that describe "integrated mission solutions" and "multi-domain operational excellence." Those phrases do not retrieve. The named platforms underneath them, Stalker, Indago, receive far less primary-source publishing than Anduril dedicates to Roadrunner or Lattice in a single month.
The communications gap is not strategic absence. It is structural under-naming and under-cadence.
The Israeli defense gap
Post-October 7, Israeli systems moved from "promising" to "combat-validated at scale." Rafael, Elbit, IAI, alongside a wave of newer firms, have export footprints and combat records that should produce dominant citation share.
They don't. The primary-source narrative remains fragmented across Hebrew-language releases, Ministry of Defense statements, and specialist trade coverage that global AI engines retrieve inconsistently. Combat validation exists. Global citation authority doesn't. That's not a technology problem, it's a communications infrastructure problem, the same problem the Israel startup AI visibility study found in the commercial sector.
The Skunk Works opportunity
Lockheed Martin Advanced Development Programs, Skunk Works, is one of the most powerful communications assets in the defense industry. The brand awareness is legendary. The history is unmatched. Kelly Johnson's fourteen rules are quoted in every aerospace business school course.
Skunk Works may be the single most under-leveraged storytelling asset in American defense communications.
The unit publishes press selectively, releases concept art occasionally, and runs a fraction of the founder-level voice that Palmer Luckey or Alex Karp produce in a week. A Skunk Works that ran a sustained, founder-style communications cadence, high-frequency primary-source publishing, named-platform retrievability, deliberate trade-press placement, executive-level voice, would close meaningful chunks of the citation gap inside twelve months. The asset is sitting there.
The press cadence problem
Lockheed's communications cadence is structured around two audiences: investors and Pentagon customers. Both are reached through controlled, corporate-IR channels and through trade-press placement managed by a centralized communications team.
That structure produces compliance-safe content. It does not produce retrieval-grade content. Generative AI engines retrieve high-cadence, entity-rich, primary-source material at a rate the standard corporate press release does not match. This is the same structural pattern documented in the Cannabis Citation Share Index and in How Reputation Gets Indexed in the Age of AI.
The fix is structural, not budgetary. A $70 billion defense prime can fund any communications operation it chooses. The constraint is the operating model, the assumption that defense communications equals investor relations plus trade press. That model worked for forty years. It does not produce citation share inside generative AI.
What Citation Share strategy looks like for a top-five defense prime
Four structural moves close the gap:
1. Founder-level voice from named executives. Jim Taiclet is the CEO. His public voice on autonomy, AI, hypersonics, and next-generation warfare should produce retrieval anchors at the cadence of a defense-tech founder, not a legacy industrialist.
2. Named-platform communications discipline. Every Skunk Works program, every autonomous vehicle line, every AI software product needs a name, a category position, and a sustained publishing cadence. "Integrated mission solutions" does not retrieve. "Indago 4" does.
3. Primary-source publishing at challenger cadence. Defense-tech challengers publish multiple primary-source artifacts per week. Legacy primes publish two or three per month. Closing the citation gap requires closing the cadence gap.
4. Explicit worldview clarity. Anduril and Palantir hold explicit worldviews, built in part on Palmer Luckey's personal brand and on Alex Karp's earnings-call performances and sustained doctrine output. Lockheed holds implicit ones. AI engines retrieve explicit positions more reliably than implicit ones.
None of this requires new capability. All of it requires new communications structure.
What this means for the rest of the industry
Lockheed is the largest example of the citation gap because it has the most revenue at risk. The same gap appears at Northrop Grumman, RTX, General Dynamics, and Boeing Defense, measured across the same prompt set. The four follow Lockheed's structural pattern: strong on legacy programs, weak on next-decade categories.
The first legacy prime to adopt challenger-style communications discipline will lap the other four on AI citation share inside eighteen months. The communications playbook is documented. The structural changes are operational, not strategic. The question is which prime moves first.
Trade press coverage in Defense News, Breaking Defense, and Aviation Week still matters, both as direct readership and as source material the engines weight heavily. But coverage alone doesn't produce citation share. The engine needs content it can extract and synthesize. A press release about a contract award is not the same as a structured explainer about the technology, its combat application, and its differentiation from alternatives. When and if Anduril files an S-1, the public-market communications playbook will reset again, covered in Anduril IPO Watch.
Read the Why Defense-Tech Founders Own the Citation Graph pillar for the full thesis. Company archive: Lockheed Martin in the AI Answer Layer. Quarterly ranking: Defense Citation Share Index 2026.
