Air Force Secretary Troy Meink announced Monday at the the Air & Space Forces Association’s 2026 Air, Space & Cyber Conference that the service has on-orbit space control weapons capable of defending the joint force against “hostile adversary action.”

Meink announced in a keynote address Monday at the Air & Space Forces Association’s 2026 Air, Space & Cyber Conference that the service has on-orbit space control weapons that are capable of defending the joint force against “hostile adversary action.”

Meink offered no other details, saying only that the phrasing was “well thought out.” He said maintaining and being able to depend on U.S. space capabilities is necessary as adversaries develop their own capabilities to threaten them.

“It is critically important that we maintain our dominance not only in the air but in space, and so that we’ve had to take steps to make sure that when we’re threatened, we can take care of that,” he said during his speech.

During a media roundtable following the keynote, Meink declined to comment on the type of weapons or why the announcement was made now.

Meink’s statement is the most direct public acknowledgment by a U.S. official to date regarding space control weapons, but he did not say when the weapons were launched or how many are in orbit.

Meink said that for more than a decade, the U.S. has been aware that adversaries such as China and Russia were developing their own capabilities to threaten U.S. systems.

“We need to maintain dominance in space as it becomes more and more important to how we function, how we operate militarily and economically, civilian-wise,” he told reporters.

The secretary called it necessary for deterrence, but he would not give more details about the weapons, saying that would negate its purpose.

“I’m not going to talk about the specifics of what we’re doing, whether we’ve tested or not tested or anything else,” he said.

During a panel regarding counter-drone defenses at the Air and Space Cyber Conference, Hiatt noted that “everyone talks about cost asymmetry… and how important that solving that cost asymmetry problem is. But for the folks that are on the ground, the battle captain trying to make that assessment, it’s not a cost asymmetry problem. It’s a magazine doubt issue.”

“When you have a problem like UAS threats, they’re coming in mass,” Hiatt continued. Trust in an airman’s system, he emphasized, is crucial when the threat time has been cut down to seconds.

“I know that my DE [Directed Energy] weapons are gonna be able to take it out, and do so in a low-collateral way, in a very low cost per shot way, and most importantly, in a way that preserves my magazine — my scarce interceptors, and then they’ll be able to push that button.”

The challenge, according to Hiatt, is “putting that work in and building that trust. It’s not science fiction. It’s physics fact.”

For Hiatt, utilizing AI in counter-drone defense is a tool with humans initially “in the loop” but ultimately working towards a system where humans start the loop and then end up on the outside of the operating system. However, building that trust amid an industry that is still formulating its own AI policy regulations is difficult, the panelists concede.

But despite this, the panelists argue, the proverbial cat is out of the bag regarding artificial intelligence, and the U.S. military must work on how to combat fast and reckless enemy AI agents that don’t necessarily operate with the same oversight.

The military must continue “to evolve, because the threat is evolving,” said panelist Colton Wood, chief technology officer of Digital Force Technologies. “And if we don’t evolve our defenses to basically counter that, we’re just gonna always be on our heels.

During the panel, “Rethinking the Stockpile: Building the Right Mix for Future Conflict” at the Air and Space Cyber Conference on Monday, Lyons noted that the service is working to balance “capability, capacity cost and the adaptability to produce at the speed and scale for future conflicts.”

Citing the conflict in Ukraine, where battlefield adaptations can come within hours and minutes, not months and years, panelist Bryon Kroger, founder and CEO of Rise8, noted that “when it comes to software, we’ve just historically been really bad. Like, in some cases, we’re building software slower than hardware. It’s wild.”

“You wouldn’t think about not establishing a refueling apparatus for aircraft that need to get across long distances,” said Kroger. “But we haven’t really thought about, from an enterprise IT perspective, how to create a similar refueling apparatus or the ability to continuously deliver software to the field in the way that you should. It’s not getting nearly enough attention and we need to start.

“We really have to decouple the software cycles from the hardware cycles. … We [need to be] deploying multiple times a day, where we can get an update to the field on demand. As soon as we need it.”

Ryan Hartman, chief executive officer of Ondas Sentinel, agreed.

“Inventory on the first day of the conflict is one important equation. But it is the ability for that inventory to evolve with the conflict. In my 27 years in the industry, I haven’t seen an adversary comply with our acquisition system yet. And therefore, we have to make sure that our acquisition system enables us to keep up with the pace of the conflict and the adaptability necessary to stay relevant in the conflict.”

When addressing how the government and industry can work together to address stockpile bottlenecks, both Hartman and Kroger offered simple, if possibly difficult to enact, solutions.

For Kroger, the biggest bottleneck Rise8 faces is something simple: a calendar.

According to him, “80-85%” of his company’s time is spent waiting on government authorization to operate — time spent “waiting in the queue” to get on the ”authorizing official’s calendar… if it takes six months to get on the ATO’s calendar, right there you’re at a minimum of six months to fill the system, but they’re going to find something because we always find things during test and evaluation. And now we go back, redevelop, and wait another six months to get on that calendar. This is how you get 12–18-month ATO cycles,” he noted.

For Hartman, the bottleneck within U.S. industry is clear: “The first is the supply chain, the second is the supply chain, the third is the supply chain and the fourth is testing.”

With technological advancements unfolding at breakneck speed amid conflicts in Eastern Europe and the Middle East, time is not a luxury the U.S. Air Force and its industry partners have much of.

That has been a boon to people like Kroger, who notes that a government that unleashes performance-based outcomes means that the defense industry is not an insider game anymore. Rather, contracts go to companies that offer solutions-based, scalable products.

As for the most important action the government and industries must take to be better positioned in the next few years, Hartman was blunt:

“Make sure we have capacity prior to a conflict.”

Secretary of the Air Force Troy Meink used the names for the first time during a Monday keynote address at the Air & Space Forces Association’s 2026 Air, Space & Cyber Conference while announcing that the Air Force intends to have at least 500 autonomous aircraft in service by 2032. The CCAs are expected to perform many of the same missions manned fighters complete currently.

“Because of the steps we are taking today, we will have a significantly different force by 2032,” Meink said at the conference. “By then, our special operators will have the ability to employ thousands of autonomous one-way attack systems, and we will have autonomous fighters like the Collaborative Combat Aircraft.”

Meink highlighted that the aircraft went from contract award to first flight in fewer than 18 months, “almost unheard of in aircraft development.”

General Atomics’ FQ-42A took its first test flight in August 2025 — over a year after its contract was awarded. Anduril, the second defense company tasked with production of the FQ-44A, rolled its first aircraft off the assembly line in July after flight testing in October 2025.

Recently, the force completed its first-ever live-fire test with Anduril’s FQ-44A, executing an end-to-end, beyond-line-of-sight strike against a simulated target using an inert AIM-120.

The CCAs are not the only class of autonomous aircraft the service is “aggressively” pursuing, Meink said, spotlighting how the ISR strike platforms have been essential during Operation Epic Fury and how the MQ-9 Reaper and the Unmanned Long-endurance Tactical Reconnaissance Aircraft, or ULTRA, have both been utilized by the service.

The Air Force is furthering the development of Massed Modular Aircraft, or MMAs, a family of low-cost, multi-role strike platforms meant to succeed the MQ-9 Reaper. Meink asserted that the service intends to field 100 in 2029 at a lower cost than the CCAs and a fraction of the cost of manned aircraft being built today.

“MMAs will provide affordable, attributable, long-range strike, and we will be able to field them at scale,” Meink said. “We expect to have tens of thousands in the inventory by 2030. Matter of fact, they’re already being used today.”

Meink said that for his goal to become a reality, Congress and industry partners need to provide more resources to generate prototypes and production at scale.

For the CCA program, the service’s fiscal 2027 budget request included $996.5 million in procurement funding, as well as $150 million in advanced procurement for the following fiscal year and $1.37 billion in continued research and development, for a total program request of $2.37 billion.

The new UAS, dubbed Gamera, comes as the Air Force searches for risk-tolerant aircraft that can be produced cheaply and in large numbers as part of its new Massed Modular Aircraft, or MMA, program.

The company revealed the platform at the Air & Space Forces Association’s annual conference near Washington.

In a press release, the company said Gamera is the first UAS designed to carry bomber-class strike munitions such as the Long-Range Anti-Ship Missile, or LRASM, and Joint Air-to-Surface Standoff Missile-Extended Range, or JASSM-ER, as well as other cruise and hypersonic missiles.

The company added that it “will be able to deliver all of the capabilities of the MQ-9, as well as employ payloads that the MQ-9 cannot, at a fraction of the price.”

Although Swarm did not give a specific price point, the company said comparable strike effects would cost 10 times less than existing platforms. The company also cited recent Defense Innovation Unit statements, calling existing systems “unsustainable.” The MQ-9 Reaper purportedly costs more than $30 million apiece.

Oliver Palmer, Swarm’s chief revenue officer and co-founder, explained that the design’s lower cost will enable the U.S. and partner nations to find other civil and commercial applications for the new UAS.

“The next few decades will see a continuous contest with China over global power projection, first focused near China over Taiwan, over time extending globally,” he said. “The [People’s Liberation Army] has built asymmetric ways to challenge our legacy force structure.”

Palmer added that “we cannot continue to pursue incremental improvements in legacy approaches and hope to protect U.S. interests and deter [the People’s Republic of China] coercion of our allies and partners.”

The Gamera is classified as a Group 5 UAS, the largest category of unmanned aircraft under the Defense Department’s classification system. The company said it has a roughly 72-foot wingspan and can fly 9,000 nautical miles before needing to refuel.

Swarm said it has invested tens of millions of dollars into Gamera’s design since it launched in 2022. Over the past four years, the company has been awarded millions in Defense Department contracts and raised more in investments.

The company described Gamera as “well positioned to support the generational transformation the Air Force is advancing with MMA.”

The Pentagon now wants artificial intelligence that can cut through the confusion.

It’s seeking “AI-based software tools to turn fragmented sensor data into a continuously updated understanding of space and missile threats fast enough to drive operational decisions,” according to the Defense Innovation Unit solicitation, which is due Sept. 24. “Existing tools struggle to distinguish closely spaced objects, track emerging threats, and keep threat models current.”

The project, known as the “Space Threat Intelligence Synthesis Engine,” will leverage AI’s greatest strength: the ability to quickly analyze reams of data. The software will fuse “massive streams of multi-source data, including live video, satellite imagery, radar and sensor feeds, geospatial data, and classified intelligence reports.”

“By automating advanced cross-data correlation, the architecture aims to uncover hidden patterns, complex operational relationships, and predictive threat behaviors far beyond the capabilities of human analysis alone.”

Ultimately, the system should “generate actionable, confidence-scored alerts supporting analyst-in-the-loop, human-on-the-loop, or automated workflows.”

The AI will help human operators — and other machines.

It will provide “intuitive, human-readable visualizations for frontline operators and low-latency, machine-to-machine APIs to drive automated command-and-control (C2) workflows.”

DIU is aiming for an open-source system with a latency delay of no more than five seconds — and preferably no more than two seconds — between the arrival of data and the results being displayed. Throughput should be 20 to 30 megabytes per minute, with bursts of up to five gigabytes.

Contractors should demonstrate that their products show “significant speed and accuracy improvements in event detection, characterization, and attribution compared to existing baseline systems.”

As skies fill with missiles and drones in conflicts like the ones with Ukraine and Iran, AI is increasingly being seen as an indispensable partner to overworked human operators. In particular, with a perpetual shortage of expensive interceptors, such as the Patriot air-defense system, operators must quickly decide when and where to allocate them.

Defense contractors and AI firms are teaming up to incorporate AI into missile defense.

For example, Northrop Grumman announced in July that it was partnering with AI firm Camgian to incorporate artificial intelligence into integrated air and missile defense systems. BAE and Scale AI have also teamed up for AI-enhanced missile defense.

The wing said the deployment, in support of the Air Force Weapons Instructor Course, produced “a historical milestone, dropping live bombs from operational F-15EXs for the first time.”

The images show jets on the ramp and in the air on Aug. 27, some carrying a live 500-pound Mk-82 on the centerline station and others carrying an AN/AAQ-33 Sniper targeting pod. The Air Force has not said how many bombs were dropped.

The unit first dropped inert bombs in May at Mountain Home Air Force Base, Idaho. The wing announced in July it dropped 24 BDU-50s in the 500-pound class and 12 BDU-56s in the 2,000-pound class, all concrete-filled practice bombs, across 10 missions.

The Mountain Home drops were the first of any kind by an operational F-15EX unit.

The 142nd flew the F-15C, an air-superiority fighter, before receiving the Air Force’s first operational F-15EX in June 2024. Three of the unit’s pilots already had air-to-ground experience and built a compressed syllabus for the rest of the wing’s pilots.

A small group of loaders with prior bomb-loading experience became certified to load bombs on F-15Es at Nellis earlier this year, then trained the loaders who went to Mountain Home.

The wing’s July statement said live-bombing at Nellis would follow as the next step toward the unit becoming air-to-ground mission capable.

“It’s about building a dynamic and versatile organization, and we are just getting started,” Col. Joshua Hovanas, commander of the 142nd Operations Group, said in a statement.

The Eagle II is rated for a 29,500-pound payload and can carry up to a dozen air-to-air missiles. The jets photographed at Nellis flew without conformal fuel tanks, which carry additional weapons stations.

The planned F-15EX fleet has grown since a fiscal 2023 budget request proposed capping the buy at 80 aircraft. In April, the service said it now plans a fleet of 267.

The movement, dubbed the North Star Readiness Campaign, responds to the push to maintain the readiness and availability of the U.S. bomber force for global operations, according to a Tuesday release.

Air Force Global Strike Command launched the campaign alongside Air Force Materiel Command and the Defense Logistics Agency to bring together operators, acquisition and sustainment experts and logistics partners.

“The North Star Readiness Campaign is our commitment to ensuring the warfighter has the nation’s bomber force ready to go at a moment’s notice,” Gen. S.L. Davis, commander of AFGSC, said in the release.

“By modernizing these incredible platforms and ensuring our maintainers have the parts and support they need, we are sending a clear message to our adversaries and allies alike that the United States’ long-range penetrating strike capability is, and will remain, second to none,” Davis continued.

The collaboration between AFGSC, AFMC and supply partners, such as DLA, is meant to break down logistical barriers and streamline the channel for parts, engineering support and depot maintenance to increase the number of B-52 and B-1 aircraft in the air, per the release.

The partnership between the three teams can remove hurdles and speed up lateral shipments to arrive in two to three days. Previously, those deliveries would take a week or more.

“Strengthening the supply chain and engineering support means maintenance is faster and less complex,” Matt Miller, executive director of the Logistics and Engineering directorate, AFGSC, said in the release. “It ensures these critical aircraft remain a credible strategic asset for decades to come.”

The North Star Readiness Campaign intends to boost the B-1’s operational readiness through sustainment efforts that focus on expediting parts available for critical structural and avionics components, while collaborating with AFMC and DLA to accelerate depot-level maintenance cycles, the release reads.

Since the campaign began, the number of “mission-impacting part problems” dropped from more than 450 to around 160, allowing units to shift focus from immediate emergencies to backorders and emerging failure trends.

A range of ongoing and new modernization programs, managed through the Office of the Assistant Secretary of the Air Force for Acquisition and AFMC, is unified through the campaign.

The announcement acknowledged that while the B-52 and B-1 are being upgraded to stay in service for longer, the B-2 Spirit will be maintained for combat effectiveness until the B-21 Raider is fully operational.

Markoff explained that calling in air support was an effective way to clear an area, but with thousands of potential targets and few available weapon systems, he constantly had to calculate whether a particular target was worth it.

“Even if that target wasn’t active, knowing that it might be and maintaining availability of that type of a weapon system on station in case it showed up, that was all math I was doing in my head,” he said, adding, “We had to be really careful about saving those munitions.”

In an interview with Military Times, Markoff, now the co-founder and chief executive of Smack Technologies, said his experience in Mosul illustrates the kind of problem his company’s artificial intelligence software, Omega, is designed to help solve.

Earlier this month, Markoff spoke with Military Times about Smack’s Alpha, the wrist-worn device for which the company recently raised $61 million. He followed up to discuss its flagship product, Omega. The company shared a 14-minute video demonstration typically reserved for general officers.

“What the demo shows is a Joint Fires workflow from long-range planning, through what we would call basically the deliberate planning cycle, through the phased execution horizon, to literally right now: What do we do in the context of that larger plan?” he said.

According to the video, a commander begins the process by uploading intelligence and planning documents to give Omega operational context. The commander can then customize that information by selecting factors, such as units and boundaries. Lastly, the system generates high-level plans and detailed workflows.

“The outputs of this workflow look like what a human would create. We’re just generating it automatically for them and then letting them adjust it … We’re just doing it faster and helping them run calculations that are complicated enough that you don’t want to have someone do it on a whiteboard,” Markoff said.

He explained that Omega’s goal is to determine not just which targets to strike, but how to sequence strikes over the course of a month or two to systematically degrade an adversarial force. He added that Omega gives commanders options, along with the pros and cons of each.

“Sometimes it takes people like a month to build that plan. Building the next one to four days takes one to four days,” he said. “And figuring out, ‘What do we do right now?’ Well, the good news is you only have a few seconds, and it only takes a few seconds.”

During the execution phase shown in the video, a commander can monitor operations in real time and adjust plans using information like new intelligence reports, inventory lists or weather forecasts.

For comparison, Smack’s demo likens Omega to Amazon’s approach to the “last-mile delivery problem,” which refers to the logistical challenge of finding the most efficient and resourceful way to get a package to its destination.

Markoff explained the reasoning behind the comparison, saying that while risk tolerances and resources can change during an operation, the overall objective may not.

“It’s not just a scheduling problem, but it’s a pairing of assets and scheduling of assets,” he said, adding that it’s “optimizing that usage in the context of what we’re trying to do long term.”

The demo depicts Omega generating a large number of actions, targets and destinations. The software is designed for campaign-level threats and for developing plans at both the operational and tactical levels. But his company is actively expanding the scope of what it offers to include coordinating humanitarian efforts.

“That has not been the use case that our contracts have focused on, but there is absolutely a demand signal and a requirement for it that we are getting from the users, and it is something that we are building into that longer term,” he said.

While Smack has already shared parts of the demo in marketing materials, the company is in the process of transitioning from prototyping to production, Markoff said. Smack currently has contracts with three U.S. military branches and expects to expand to all six by the end of the year.

The Department of Defense wants to take that idea one step further and wield AI-generated “personalities” to mimic real people — and fool enemy AI trying to identify patterns in friendly cellphone traffic.

The goal is to mask “digital exhaust,” the metadata automatically emitted by mobile phones, such as location and specific device IDs. Intelligence agencies, law enforcement, militaries and even criminals have become adept at using this digital debris to locate users and analyze their behavior.

In particular, “accessing foreign mobile phone networks exposes individuals and organizations to surveillance and espionage threats,” warned the DoD Small Business Innovation Research solicitation, which opens Sept. 23 and closes Oct. 21.

“Adversaries can leverage this data to track daily routines, identify deviations indicating special events (e.g., VIP visits or mission preparation), and map sensitive facilities based on cell phone concentration or periods of inactivity,” the solicitation explained.

The solicitation doesn’t specifically mention AI-created people chatting on their mobiles. But DoD does envision a “digital exhaust deception system” in which AI “personas” generate cellphone signals that mimic the behavior of real people.

Each device should have a minimum of two personas, and eventually four or more. The device should weigh fewer than 2 pounds and have a battery life of up to 36 hours.

Merely generating fake cellphone signals isn’t enough. The device must “mimic realistic patterns of life to defeat automated network analysis,” DoD said.

At a minimum, this means defeating basic traffic analysis. Ultimately, the Pentagon wants a system that can defeat behavioral tracking by advanced AI.

Phase I of the project involves creating a design concept and algorithms, as well as listing any “proprietary software required to conduct AI-enabled mission planning and operations.” Phase II will generate a prototype, followed by Phase III and delivery of “fully integrated mission management software and hardened decoy box form factors (e.g., Wi-Fi router-sized team devices and puck-sized personal devices).”

The technology could lend itself to civilian applications. A capability to obscure cell phone metadata “directly translates to protecting high-value corporate assets, personnel, and intellectual property from commercial espionage and tracking,” the solicitation states.

Analyzing signal traffic has long been a useful intelligence tool. For example, in World War II, the U.S. Navy used it to derive intelligence on Japanese naval deployment. Today, police have software that uses cell phone data to trace a user’s past movement for months.

On the other hand, traffic analysis can boomerang on its creators. Japan used decoy radio signals in 1941 to convince U.S. analysts that the Japanese fleet was at home, rather than steaming toward Pearl Harbor, Hawaii.