AMPulse

Q2 2026 news-flow analysis

The Procurement Quarter

In Q2 2026, defense overtook aerospace as AMPulse's largest named AM end market. The deeper shift was from qualification awards into capacity and deployable production systems.

Defense share of AM news flow · monthly, H1 2026

9.3

11.3

6.3

17.9

13.7

15.0

JanFebMarAprMayJun

Q1 monthly average

9.0%

Q2 monthly average

15.5%

AMPULSE DATA REPORT · Q2 2026Data snapshot: Jul 10, 2026

Executive summary

• Defense took the top named end-market rank. It moved from 9.6% of covered companies in Q1, behind aerospace at 13.1%, to 15.5% in Q2, ahead of aerospace at 12.1%.
• Customer-program coverage broadened more than financing coverage. The share of defense-covered companies appearing in partnership or contract stories rose from 30.7% to 39.4%, while funding stayed nearly flat at 16% to 16.3%.
• The more durable signal is a new procurement unit. Across engines, naval sustainment, drones and field repair, the object being bought is increasingly a qualified production system: hardware, process data, secure files, training and a deployable operating model.

Defense took the top end-market rank

AMPulse tags every article in its daily AM news pipeline with a demand vertical. In Q1, aerospace was the largest named end market by unique covered companies. In Q2, defense moved into first place: 160 companies, or 15.5% of all covered companies, against 125 aerospace companies, or 12.1%.

The monthly series shows when the rank changed. Defense-tagged stories held a 6.3-11.3% band through Q1. In April the share jumped to 17.9% and stayed high, 13.7% in May and 15.0% in June. Across the quarter that is 250 defense-tagged stories against 101 in Q1.

Story counts overstate momentum when several outlets cover the same event, so the headline uses companies. The quarter was not a handful of large deals covered repeatedly: 138 of the 160 Q2 defense-covered companies had no defense-tagged coverage in Q1. But the reverse is also important. Only 22 of Q1's 75 companies appeared again in Q2 defense coverage, a 29.3% retention rate. This is evidence of a broad quarter, not yet evidence of a stable supplier cohort.

Two collection artifacts could have manufactured this pattern, so we checked both before writing it up. First, a March 19 redesign of the pipeline's search queries could have tilted collection toward defense; it did the opposite, removing the one defense-specific query the pipeline had, and the current query set contains no defense or military keywords. The surge arrived through general AM channels. Second, overall pipeline volume rose 37% quarter over quarter as query coverage expanded, which is why every figure above is a share, not a count. Native-script and Latin duplicates of the same company (two cases in the defense set) were merged before counting.

Customer programs broadened faster than funding

The normalized category mix is narrower than the raw story counts suggest. Partnership and contract coverage expanded from 23 to 63 unique companies, increasing from 30.7% to 39.4% of the defense cohort. Funding coverage rose in count but stayed flat in mix, 16% to 16.3%. Expansion and capex also grew in company count, 14 to 27, but slipped from 18.7% to 16.9% of the much larger Q2 cohort. The category evidence supports a shift toward customer programs; it does not show every production category gaining share.

The curated ledger gives the shift a harder edge. It contains $69.1M in disclosed procurement contracts and qualification program awards, plus $57.3M in supplier equipment and system commitments. Together they total $126.4M across six events. Firestorm Labs' $82M Series B remains the quarter's largest single defense-AM venture event, but this report does not claim to measure all venture funding. The comparison is stage evidence, not a capital-flow census: public customers funded qualification while suppliers committed capital to production capacity.

The procurement unit became a production system

Inside the quarter sits a tighter pattern: the production cell that ships to the point of need. Five Q2 events, from four countries, describe variations of the same product category.

Firestorm Labs raised its $82M Series B specifically to scale xCell, a containerized manufacturing platform the company positions for building drones near the point of use. AML3D commissioned its first containerized ARCEMY metal printer at the US Navy's Additive Manufacturing Center of Excellence in Danville, Virginia, redeployable in days rather than the weeks a fixed installation needs. Omni3D and WITU built their Mosquito munition scaling plan around the TFU20, a power-autonomous containerized print cell. In Korea, LinkSolution demonstrated its AM Fab mobile system at an Army drone pilot unit in May and extended a field-production MOU with the 7th Logistics Support Group in June. Austal USA launched Digital SEA, a Navy-hosted platform for moving qualified AM part data securely to wherever the printer happens to be.

The common architecture is more important than the container. Defense buyers are assembling five layers: production hardware, a qualified part family, material and process data, a secure digital thread, and an operating model that includes power, logistics and trained personnel. In that frame, the factory's location becomes part of the mission system. Q2 is the strongest current cluster in AMPulse's news flow for that architecture moving from pitch to field program, platform launch and equipment order.

What pulled the quarter forward

Nothing about Q2 was spontaneous. The quarter is consistent with a pipeline that had been filling for two to three years: budgets funded qualification, qualification opened contract vehicles, and suppliers began adding capacity. Four mechanisms help explain the timing, but they do not carry equal causal weight.

Budget cycles reached the contract stage. The US FY2026 budget request contains about $3.3B in additive-manufacturing-related projects across 16 programs, an 83% increase over the prior year, per an analysis of DoD budget justification documents by trade outlet 3D Printing Industry. Request-level figures are not appropriations. The Q2 evidence at the other end is concrete: a $31M NAVAIR qualification contract to Colibrium Additive, a $29.7M Air Force engine-qualification contract to Beehive Industries, and an $8.4M AFRL titanium-aerostructures program with GKN Aerospace.

Qualification programs converted into supplier commitments. The Defense Logistics Agency selected 24 companies for its JAMA IV pilot parts program in late February; award coverage, including Nikon AM Synergy's Long Beach selection, landed through April. Beehive's June order of 30 EOS metal systems is the clearest capacity signal: a defense supplier adding series-production equipment while its engine family moves through flight qualification. The source announcements establish the sequence, not the eventual utilization of those machines.

Regulation set a future eligibility boundary. Section 849 of the FY2026 NDAA, enacted December 18, 2025, bars the Department of Defense from operating or procuring AM systems made in, or networked through, China, Russia, Iran or North Korea, with the prohibition taking effect one year after enactment. It gives domestic and allied suppliers a clearer eligibility channel from December 2026. This report does not attribute any specific Q2 order to the clause.

European rearmament created supportive context, not a traced funding chain. SIPRI's April 2026 release put 2025 European military spending at $864B, up 14%, with Germany up 24%. The EU's EUR 150B SAFE loan instrument completed its national-plan approvals on April 10, with Poland holding the largest indicative allocation at EUR 43.7B. The quarter also contained AnyShape's Eurodrone production selection, Omni3D's work with Poland's Military Institute of Armament Technology and a follow-on E-PBF order from an unnamed Swedish defense customer. None of those source releases ties the event directly to SAFE, so the instrument belongs in the macro context rather than the quantified ledger.

Attritable-drone procurement raised the value of distributed production. The US Army's low-cost drone production "gauntlet" runs its first phase from February to July 2026, targeting 30,000 drones from 12 vendors. The program is a demand signal for low-cost, rapidly replenished systems; it is not itself an AM contract. Its relevance is operational: the higher the replacement tempo, the more valuable qualified near-point-of-use manufacturing and repair become.

Where it happened

The United States anchors the quarter with 137 stories across 90 companies, consistent with the budget and NDAA mechanisms above. The next tier is more interesting. The United Kingdom (24 stories) rides US programs as much as domestic ones: GKN's AFRL award and Q5D's wire-harness cells for the US Army both sell British capability into American demand. Australia (21 stories) is almost entirely the AML3D-US Navy relationship. Poland, Sweden and Belgium are small in count but each carries a European-rearmament story with a named program behind it.

South Korea's 22 stories across 12 companies deserve their own paragraph, because the shape is different. The targeted Korean-language searches used for this report surfaced no new DAPA or ADD additive initiative announced between April and June; that is a search result, not a claim that no program exists. The visible activity is bottom-up: LinkSolution's two army-unit engagements, and the ongoing build-out of Daejeon Technopark's defense-and-space AM joint manufacturing center, a MOTIE-funded program placing eight large AM systems by 2027. Korean defense AM in Q2 was carried by a KOSDAQ-listed printer maker and regional infrastructure more visibly than by a new acquisition-agency announcement. Whether that changes in Q3, as field-unit MOUs mature into supply contracts, is one of the specific things this dataset can watch.

The verified ledger

The table below is the curated core of the quarter: 19 events that survived a primary-source check. Every amount was verified against a company release, SEC filing or government announcement; events that only exist in trade coverage are marked accordingly. Amounts are announcement values, not obligations. Six production-side events make up the $126.4M total: $69.1M in procurement and program awards, and $57.3M in equipment and system commitments. Venture rounds, the Markforged acquisition and undisclosed-value events are listed but never summed.

Date	Company	Event	USD M	Source
2026-04-06	Austal USAUS	Platform launch *	n/d	source
2026-04-09	Beehive IndustriesUS	Procurement contract	29.7	source
2026-04-13	GKN AerospaceGB	Program award	8.4	source
2026-04-15	Nikon AM SynergyUS	Procurement contract *	n/d	source
2026-04-21	FreemeltSE	System order *	n/d	source
2026-04-22	Colibrium AdditiveUS	Procurement contract	31.0	source
2026-04-29	Firestorm LabsUS	Venture funding *	82.0	source
2026-05-06	AML3DAU	System order	0.9	source
2026-05-21	LinkSolutionKR	MOU / field program *	n/d	source
2026-05-26	AnyShapeBE	Production partnership *	n/d	source
2026-05-27	StratasysUS	Acquisition *	42.5	source
2026-06-09	Omni3DPL	Production partnership *	n/d	source
2026-06-16	Beehive IndustriesUS	Equipment capex	50.0	source
2026-06-16	Foundation AlloyUS	Venture funding *	22.0	source
2026-06-17	Divergent TechnologiesUS	Capacity expansion *	n/d	source
2026-06-17	Q5D TechnologiesGB	Production partnership *	n/d	source
2026-06-19	AML3DAU	System order	6.4	source
2026-06-22	6K AdditiveUS	Capacity expansion *	n/d	source
2026-06-25	LinkSolutionKR	MOU / field program *	n/d	source

* Not counted in the quantified production-side total: venture capital, consolidation, undisclosed values, and AM-adjacent events. Company names link to their AMPulse profiles; "source" links to the primary or best-available announcement, with full source lists per event in the report data.

One correction from the verification pass is worth recording publicly. The Colibrium NAVAIR contract circulated in one defense outlet as $3M; the GE Aerospace release and two trade outlets state $31M, and this report uses $31M. Three widely shared events fell outside the window and are excluded from all Q2 counts: Velo3D's $9.8M DLA award (March 30), AML3D's A$2.61M submarine-parts order (March 26), and Deloitte's investment in drone maker KIHOMAC (October 2025), which resurfaced in May coverage and reads like a Q2 event until you check the date.

What this signals

Limits

News flow is a proxy for procurement, not a measurement of it. AMPulse counts stories and companies; it does not yet reconcile them against obligation data such as USAspending. A contract announced is not a contract obligated, and announcement values (including IDIQ ceilings) routinely differ from eventual spend. The JAMA IV case shows the failure mode inside the sources themselves: trade coverage describes a roughly $9.8M collective ceiling across 24 companies, while one participant announced the same figure as its own award ceiling; this report treats the program total as approximately $9.8-10M and does not sum per-company claims.

The vertical tag is assigned by the enrichment model at collection time. This report accepts the news pipeline's existing AM screening and does not re-review every article for AM relevance. The artifact checks in the first section address collection bias and duplicate coverage, not classification error. Stored rows also do not carry classifier-version metadata, so the Q1-Q2 comparison assumes the vertical tags are comparable across the half-year. The named-end-market ranking and 19-event ledger provide two different evidence layers, but only the ledger was independently source-verified event by event.

Coverage skews toward English-language sources despite CJK-specific collection channels, so the US share is likely overstated relative to China's actual defense-AM activity; the three Chinese defense-tagged stories in Q2 should be read as a floor set by observability, not a measurement of Chinese activity. Quarter boundaries are story dates and cluster around trade events such as Sea-Air-Space, RAPID and Eurosatory. The report does not separately quantify that calendar effect, so part of the Q2 step may reflect when the industry chose to announce programs.

Three observations summarize the quarter. Defense overtook aerospace as AMPulse's largest named end market. Partnership and contract coverage widened its share of the defense cohort while funding stayed flat. The curated ledger then showed $69.1M in public awards alongside $57.3M in supplier capacity and system commitments. Q2 2026 is the quarter the defense pivot became visible as a procurement system. Whether it becomes a durable demand anchor is now a retention, obligation and utilization question.

Lee, S. (2026). The Procurement Quarter. AMPulse Data Reports. https://www.ampulse.online/reports/q2-2026-defense-procurement