Defense One - Science & Tech

GenAI.mil saw more than 2 million users in one week, top DOD official says

Edward Graham — Wed, 23 Sep 2026 22:00:00 -0400

The Pentagon’s enterprise-wide generative artificial intelligence platform is continuing to see high levels of use across the department, with millions of civilians, contractors and military personnel taking advantage of the service.

In a discussion with Punchbowl News on Wednesday, Defense Department Chief Digital and Artificial Intelligence Officer Cameron Stanley said “as of last week, we've had over 2 million people in a week across all three models log in and use AI.”

GenAI.mil, which DOD launched in December, currently hosts three AI models: Google Cloud's Gemini for Government, OpenAI’s ChatGPT Mil and xAI’s Grok for Government. The service provides vetted, internal AI tools to help users with unclassified tasks, such as analyzing information and creating and refining documents.

The department employees roughly 3 million total personnel who can access GenAI.mil.

Stanley called the platform “a huge, a huge win for the department,” adding that he initially thought AI use across the Pentagon “was going to be a harder sell.”

“Before GenAI.mil was rolled out, we had a total of maybe 80,000 folks that were using AI across the entire department in similar deployments,” he said. “The first week that GenAI.mil was released last year, we had over 300,000 people log on.”

Stanley said users were also quick to take advantage of Google Gemini for Government’s Agent Designer tool, which was rolled out on the platform in March and allows approved civilians, contractors and military personnel to create custom AI agents that can help them automate rote tasks.

“Within two weeks [of its deployment], we had over 50,000 agents that people are using,” he said. These aren't complex, robust agents; they're helping people with their daily workflows.”

The latest figures continue to show a rising uptick in DOD use of tools available through GenAI.mil. During an event in July, Stanley said the platform had reached a then-record of 1.7 million users, who had created more than 100,000 custom agents.

Wednesday’s discussion about the Pentagon’s uses of AI comes amid growing concerns from some lawmakers and tech executives about the dangers posed by advanced models, as well as how the capabilities will be utilized moving forward. During a speech at the UN on Tuesday, President Donald Trump voiced his continued opposition to slowing AI development, saying “we’re going to encourage it, not rein it in.”

Stanley similarly argued that “unilateral disarmament is the wrong approach.”

]]>

'Super Intelligence': the president’s new term for AI, explained

John Croxton — Tue, 22 Sep 2026 21:16:27 -0400

President Donald Trump on Tuesday announced a rebranding of the term “artificial intelligence,” saying during the United Nations General Assembly that from now on, it will be called “super intelligence.”

“From this point forward, all of United States’ documents, and hopefully the world’s, will be changed to use the much more accurate term, ‘super,’ as opposed to ‘artificial’,” he said.

Though the rebrand may have been a surprise to many, superintelligence is already an established term in AI literature. It was popularized in philosopher Nick Bostrom’s 2014 bestseller Superintelligence, which defines the term as “any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest.” Artificial superintelligence, or ASI, thus refers to future AI systems that could be much smarter than humans, and contrasts with the weaker forms of AI that exist today.

ASI is a step beyond AGI, or artificial general intelligence, which OpenAI defines as “highly autonomous systems that outperform humans at most economically valuable work.” By this definition, today’s AI isn’t AGI yet, much less ASI, since we can’t yet hire AI as a reliable virtual employee.

Some experts dislike the term superintelligence, or question whether it’s a coherent concept. In the influential 2025 paper “AI as Normal Technology,” authors Sayash Kapoor and Arvind Narayanan write, “We do not think there is a useful sense of the term ‘intelligence’ in which AI is more intelligent than people acting with the help of AI.”

But others say the possible renaming contributes to an ongoing conversation about AI’s growing abilities.

“‘Super’ does align well with the fact that AI is exceeding human capabilities in a growing range of tasks and by increasing margins, which can have distinct national security implications beyond simply matching human performance,” said Caleb Withers, a research associate for the Technology and Security Program at the Center for a New American Security.

For example, in early September, OpenAI announced a solution to the Navier-Stokes Millennium Prize problem, one of the most important open problems in math.

As AI becomes more capable, it has grown increasingly central to the U.S.-China relationship. Trump and Chinese president Xi Jinping are slated to meet this week to discuss trade and AI issues. Bostrom, who has warned of the dangers of fully superintelligent AI systems, said an all-out AI race increases the risk that the technology gets out of control. “If competitors are convinced that coming second [in the AI race] means the total loss of everything they care about, they will take whatever risk necessary to bypass their rivals,” he said.

Trump reportedly worries that an AI slowdown could cause a market crash, and sees “super intelligence” as a critical sector for America to stay ahead of China. “Whoever wins AI, you have to remember this, and now I say, whoever wins SI, whoever wins super intelligence, wins. That’s the group that wins.”

]]>

NATO’s thumbs-up for AWS shows Europe still needs America

Patrick Tucker — Tue, 22 Sep 2026 11:47:35 -0400

Amazon Web Services has become the first cloud provider cleared to handle sensitive information for all NATO members, underscoring that an increasingly independent Europe still relies on capabilities made in America, and suggesting that it's not as hard for U.S. firms to compete there as the White House insists.

The company announced on Tuesday that it has received alliance approval to handle information at the NATO RESTRICTED level, which is not classified but requires safeguarding.

As European militaries fuse radar, satellite, and even acoustic information to more easily detect Russian drones and other threats, NATO-approved rapid data and information sharing is critical.

AWS, which already handles sensitive military and intelligence data for many NATO members, beat its competitors—primarily from other U.S. cloud providers such as Microsoft—to win the alliance’s first blanket approval. That’s the result of “a sustained, multi-year effort,” David Appel, acting vice president of Worldwide Public Sector at AWS, said in the company’s Tuesday statement, AWS worked with Spain’s National Cryptographic Centre, or CCN, to test its services against NATO security directives; CCN then shared its findings to all member states.

The announcement also reflects a broad, evolving NATO digital transformation strategy. The alliance is eager to incorporate more commercial information technology into its coordination and planning. And AWS remains the biggest commercial cloud operator in Europe, with nine data-center clusters, or regions, across Europe, including the U.K.

“Strengthening NATO’s ability to securely leverage commercial technology is key to building a more resilient and agile Alliance,” according to Dylan Browne, general manager of the NATO Communications and Information Agency, or NCIA, as quoted in Amazon’s statement.

While NATO members and AWS cloud users across Europe will be able to share their data more easily, they will also be able to retain control over it. The 2024 European Union Data Act requires cloud providers to protect EU data from other governments, including the United States. Practically speaking, that means housing data on European soil beyond the reach of U.S. court orders, although that’s not an explicit requirement of the Act. AWS, like Microsoft and other enterprise cloud providers, has been investing in European versions of its cloud stacks to meet those Data Act requirements.

Tuesday’s announcement shows both that U.S. technology is still a core feature of NATO military capability and that it is central to NATO’s future plans. But it also shows that European Union regulations on data storage aren’t a challenge to the success of U.S. cloud companies across Europe, even though White House rhetoric and executive orders suggest otherwise.

]]>

Will airmen trust AI? The Air Force’s future plans depend on it

Patrick Tucker — Sun, 20 Sep 2026 08:00:00 -0400

NATIONAL HARBOR, Md.—As Air Force leaders prepare to hand AI more decision-making power in the years ahead, they're also working to get airmen comfortable with that.

“A little bit scary, but we need to lean into it,” Air Force Secretary Troy Meink said at the Air and Space Association’s Air, Space & Cyber conference this week. “Autonomy is one of, if not the key technology enabling cost-effective combat power.”

Although Meink didn’t say so explicitly, “leaning in” will require new thinking on building trust between human airmen and the AI assistants, chatbots, and autonomous drones that will surround them. But other parts of the Air Force have spent a lot of time thinking about it, especially as the service works to field 500 highly-autonomous collaborative combat aircraft, or CCAs, by 2032.

“There’s a trust and autonomy question,” said Michael Gregg, director of aerospace systems at the Air Force Research Laboratory. “We have psychologists and behavioral scientists in our 711th Human Performance Wing that are actually studying this problem: How do you actually team machines with humans to make the human really really good at what they do and aid their decision-making? That is the future.”

And for some missions, like protecting air bases from drone attacks, the Air Force is exploring the possibility of not augmenting but replacing human decision-making. The waves of drones striking Ukrainian cities and ravaging U.S. bases in the Middle East have service leaders exploring the use of agents—AI tools that can take action—to work the process of obtaining approvals to launch counter-attacks.

That will require humans to define acceptable risk for allowing AI agents to make decisions.

“There is an increased push on how we use agents and agentic workflow into everything from fires to our back-office activities, and like you'll see a lot more of that coming this year,” said Ben Van Roo, the CEO of Legion Intelligence, a company that develops agentic AI agents for national-security purposes.

The question of AI directing “fires” is controversial, as the term suggests allowing a software program to launch a missile, fire a gun, or launch a deadly weapon. While there are some exceptions, military doctrine states a strong preference for meaningful human control over potentially lethal effects.

But cyber, directed energy, and other non-lethal technologies are giving the military more fires it might deploy without human oversight, said Mike Hiatt, the chief technology officer at Epirus, a company that makes microwave and other directed-energy counter-drone defenses.

“If you can have an option that has lower collateral effects and no collateral damage, you know, which is what the directed-energy weapons offer, you start to be able to change your risk calculation. You start to be able to make decisions where you say ‘OK, I'm a little more comfortable with putting this part of the system on a fully autonomous mode,” Hiatt said.

Van Roo said the decision processes to allow autonomous cyberattacks or nonlethal effects could eventually be accelerated with AI, or even automated completely.

“We're really still in the infancy right now of how we think about where we're going to use agents, what are we going to allow them to decide on? How do they work their way into our tactics, techniques, and procedures, into our doctrine?” he said.

Getting leaders and operators to be comfortable with AI agents will require new doctrine, training, and permissions, as well as a high degree of confidence that AI tools will perform in predictable ways.

“This is going to happen right now in front of us before AI and killer robots take over the world,” he said.

This year’s conference occurred amid a national debate over the dangers of AI, how it should be regulated, and who should decide.

The summer’s disclosures by Anthropic and OpenAI of surprising and dangerous model behavior were followed on Sept. 9 by the high-profile resignation of an Anthropic engineer who claimed a 10-percent probability that AI “could kill all humans.” The following day, representatives of those companies and Google intensified efforts to create an industry-safety body while signaling openness to Congressional action. But regulation is opposed by President Donald Trump, who posted this week, “The only control or ‘guardrail’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!”

Pentagon leaders’ willingness to accept AI may stem from the fact that their current systems failed to keep key U.S. bases safe. But it may also reflect several years of Defense Department efforts to grapple with concerns about intent, alignment, and safe use that frontier AI labs are just now grappling with. Since 2019, the military has had a list of ethical principles to guide AI development, experimentation and use. It also reviews processes, experiments, and testing procedures for new technology, including software and, of course, the most dangerous weapons on Earth.

But the pace of AI advancement is quickly changing concepts of how to develop and test weapons. Ryan Tseng, president and co-founder of Shield AI, told the crowd at AFA that his company had worked with Ukrainians to improve the strike capability of one of the company’s drone weapons by 70 percent through “advanced autonomous behaviors.”

If autonomous drones are the signature feature of the war in Ukraine, many military officials expect the next conflict to be defined by rapid collaboration between agents, humans, and drones that will speed up strikes and counter-strikes. That will require continuous evolution on building trust between humans and AI entities as well.

“When we think about a future that's going to have very large-scale deployments of systems that need to interoperate with each other,” said Tseng. “It's a much larger decision space. A lot more things can happen.”

]]>

Air Force contemplates unmanned tankers

Lauren C. Williams — Thu, 17 Sep 2026 19:20:29 -0400

NATIONAL HARBOR, Md.—With an eye on the Navy’s progress toward an unmanned aerial tanker, Air Force planners are asking questions about their own future tanker fleet.

“Thinking beyond the current KC-46 buy and the program extension. Whatever that next tanker looks like—our advanced tanker system—it's got to be connected. It's got to be able to defend itself. But it begs the question: Does it need to be manned? Could it be optionally manned, or could it be autonomous, or is there a family of solutions maybe that would answer that question?” Lt. Gen. Daniel Tulley, who leads Air Mobility Command, told reporters at the Air, Space, & Cyber 2026 conference.

The Air Force is currently buying Boeing KC-46 Pegasus tankers and spent $12 million to explore buying a new tanker this year. But it zeroed out funding to continue those explorations in the 2027 budget request and is asking instead for $13 million for “advanced tanker systems.”

Tulley said the Navy’s work on the Boeing MQ-25 Stingray suggests that his own service might eventually not need manned tankers. The Navy announced on Thursday that the MQ-25, which first flew earlier this year, has received approval to enter low-rate production. The service also ordered three more aircraft, bringing its total to 14.

“I do not believe that it has to be manned, a future air-refueling solution,” he said. “But then, how do we, the U.S. Air Force, bring that to bear in scope, scale, and range?”

A drone tanker for the Air Force would need to be a very large aircraft with the capacity for fuel and the ability to fly for a long time. The KC-46, for example, can fly nearly 7,400 miles and carry more than a thousand tons of fuel. The MQ-25 can fly about 500 miles and carry up to 7.5 tons of fuel.

“An optionally manned or an unmanned tanker at the ranges we need it to operate at is a very large aircraft. And we're not yet to where we're employing fully autonomous large aircraft,” Tulley said.

“Just about any large aircraft these days made by Boeing can take itself off with very little help, can fly on a programmed route and pretty much land itself on autopilot, bring itself to a stop on the runway. But we don't employ it that way with a system that's controlling it and then taxiing it in and making it a part of a more effective, broader system.”

One maker of autonomous-aircraft control systems said the vision of autonomy for large aircraft was getting clearer. And while the technology isn’t there yet, there’s progress, said Matthew George, CEO of Merlin Labs.

“We’re the furthest along in the world on making that real—in our opinion,” said Matthew George, CEO of Merlin Labs. “So we're going to be implementing autonomy on the C-130J, which is the same general size class” as the Boeing KC-135 Stratotanker.

“And the advantage with those aircraft is you can start to implement autonomy still with human crews aboard, and you can supplement those human crews in a way that's pretty helpful. That's a way that you can start to chip away at that without actually going fully to the unmanned tanker right away,” George said. “The goal is that the unmanned tanker is ready, and the autonomy ends up being ready at the same time.”

George agreed with Tulley’s point that many aircraft can somewhat fly themselves, but there are instances, such as emergency situations, when a human needs to make a decision.

“What's really hard is those edge cases and those emergencies—or those things where there are decisions to be made. So, for instance, if you're going to a destination and then there's weather in between, how do you get around the weather?” he said.

And beyond technology, there are policy challenges.

“None of this matters unless it's certified and safe. So for us, getting certification done is important, and that's what we're working really closely with the Air Force on this year,” George said. “We're taking that [autonomous pilot] brain and putting it onto the C-130J. And as far as we know, that's going to be the first time autonomy fully takes control of an operational U.S. Air Force aircraft.”

No date was provided for when that flight will occur, but the company’s technology completed a critical design review, George said.

]]>

NGA official takes top AI job at Cyber Command

Edward Graham — Tue, 15 Sep 2026 18:00:00 -0400

U.S. Cyber Command, the Pentagon’s combatant command responsible for planning and conducting military cyberspace operations, has a new artificial intelligence lead, a spokesperson confirmed on Tuesday.

“U.S. Cyber Command welcomes Rear Adm. Ronzelle Green as the Command’s new Chief Artificial Intelligence Officer and thanks Brig. Gen. Reid Novotny for his leadership and service,” a representative for the combatant command told Nextgov/FCW.

Green, who has worked at the National Geospatial-Intelligence Agency for roughly three years — most recently as its director of geospatial intelligence innovation and research — is Cyber Command’s second-ever AI head. Novotny was the first official tapped for the position last November and served in a dual role as CAIO and as cyber force generation lead for the Pentagon.

The Record first reported the news of Green’s new role.

In a LinkedIn post last year announcing his new position, Novotny said “part of my role is embracing AI because like anything new, it is not the technology but yet the culture that is difficult to change.”

Novotny further outlined his approach to adopting AI during an April conversation with Axios, saying that he was solely focused on providing Cyber Command’s personnel with the best AI capabilities, regardless of their origin.

"To survive anywhere, just in case our operators want an open-source, made-in-China model or something very boutique, we have to create the infrastructure and that ability to be agile — no politics,” he said at the time.

In addition to his prior work at NGA, Green also brings more than 30 years of IT experience in the private sector and within intelligence community agencies to his new role as Cyber Command’s CAIO. He previously served as the Defense Counterintelligence and Security Agency’s chief information officer from 2020 until 2022, and has held roles within the Pentagon and National Counterterrorism Center throughout his career. Green is also a senior officer with the U.S. Coast Guard Reserve.

Cyber Command is led by Gen. Joshua Rudd, who also serves as head of the National Security Agency in a dual-hatted leadership role. NSA and Cyber Command, however, have taken somewhat differing approaches in their adoption of and focus on AI.

Under Rudd’s watch, NSA is in the process of a major reorganization that would establish five mission centers focused on AI, China, cybersecurity, combat support and global intelligence. The planned creation of the AI-focused center comes after NSA officials have voiced support in recent months for the use of more AI tools to bolster the agency's cyber and intelligence analysis operations.

]]>

Russia is weaponizing US-built AI to make killer drones, cyberattack bots, and fake news

Patrick Tucker — Fri, 11 Sep 2026 18:53:22 -0400

Kremlin-backed actors used Anthropic’s Claude AI to program autonomous lethal drones, improve hacking and influence operations, and avoid detection, the company said Thursday.

The report from the U.S.-based AI company is only the latest warning from firms, think tanks, and others that Russia, China, Iran and others are using made-in-America tools to attack targets around the world.

Anthropic reports that a Russian “freelance” group known as GTG-27005 used Claude to build an AI model small enough to work on single-board computers but powerful enough to enable a drone to “select targets (including a ‘person’ target class) and issue detonation commands without a human in the loop.” The group also used Claude to create software to allow autonomous drones to communicate with one another to improve targeting, Anthropic’s report says. While the Russian group had not yet deployed the model to the field, they were “using real hardware-in-loop testing within their sessions,” according to the report. The company banned the group, incorporated the findings of the investigation into its safeguards, and shared details with industry groups.

Anthropic also said it had spotted Chinese and Yemeni groups trying to use Claude to design conventional weapons, efforts that Anthropic said it disrupted.

The report also detailed how foreign groups were using Claude in elaborate data theft and disinformation campaigns. It said its tools had been used to dig up information about targets, find vulnerabilities in their online accounts, create malware to bypass security and steal data, cover their tracks, and spread disinformation.

“We’ve seen groups of actors use Claude to build networks of fake social media profiles and entire news sites,” the report said.

The campaigns bear the hallmarks of Russian influence campaigns, such as the use of fake news sites and journalist profiles to spread lies about NATO activity in the Baltics. But the U.S.-built AI tools helped streamline the process, boosting the amount of disinformation and creating fake reporters that were much more believable to local audiences. Russian-linked groups “automated much of their operations from development, infrastructure acquisition, phishing, persistence through command and control, to data exfiltration,” the report said. The group even used AI to forge documents from the Central African Republic Gendarmerie and Ministry of Defense.

Another group, GTG-84005, used AI to steal census and public information that allowed its agents to tailor their messaging campaigns to specific audiences in Malaysia, where the Turkish group “laundered Russian and Chinese state media as independent reporting.”

Thursday’s Anthropic report comes two weeks after rival frontier lab OpenAI revealed that Russian actors used ChatGPT to create a fake “expert community”—the “International Burke Institute”—to push talking points to international audiences.

While Anthropic and OpenAI were able to disrupt the schemes they discovered, such operations are likely to continue as Russia uses AI to “scale production of content that aligns with Kremlin narratives,” according to an August report from the Institute for the Study of War, or ISW.

Russia is also using young people’s curiosity about AI to grow the number of humans spreading Russian disinformation, offering young people around the world “AI media internships and training” through university donations and related programs.

Where is the US pushback?

A recent article from New America notes that AI is giving authoritarian regimes a powerful new tool to distort global understanding of geopolitical events.

“Authoritarian regimes are early adopters in using AI to bend the arc of reality in their favor,” write Daniel Kimmage, the former acting coordinator and principal deputy coordinator of the State Department’s Global Engagement Center, and Darjan Vujica.

It’s the kind of thing the U.S government used to establish and run organizations to push back on, they note. But such efforts have been undermined or brought to a halt under the second Trump administration. Less than a month after taking office, Attorney General Pam Bondi dissolved the FBI’s Foreign Influence Task Force, the unit dedicated to investigating foreign disinformation and influence campaigns. Two months later, Secretary of State Marco Rubio shuttered the Counter Foreign Information Manipulation and Interference hub, known as R/Fimi.

And: “The 2025 White House AI Action Plan calls explicitly for removing any references to misinformation,” note Kimmage and Vujica.

At least when it’s made by foreigners.

“The Trump administration has enthusiastically embraced AI-generated imagery in politically divisive posts,” they write.

]]>

‘Digital employees’ are coming to the Defense Logistics Agency

Frank Konkel — Fri, 11 Sep 2026 15:32:02 -0400

The Defense Logistics Agency is piloting AI agents that work autonomously—bots that its CIO calls “digital employees.”

“This is a new frontier for everyone,” DLA chief information officer Adarryl Roberts told reporters at DLA’s Industry Collider Day on Sept. 9 in Alexandria, Va.

The move builds on DLA’s increasing use of robotic process automation, or RPA, and the generative AI tools available to its 25,000 or so military and civilian personnel. In 2025, bots saved DLA an estimated 300,000 hours of work—some working around the clock to help the combat support agency get everything from food and fuel to medical supplies and clothing where it needs to go.

“We have approximately 185 to 190 bots that are running. I’d say about 90% to 95% of those are unattended bots,” Roberts said. “So, we really cracked the nut to have those unattended. Now, we’re moving from the bots to, ‘How can we have agentic agents working across [the department], based upon our data? So, we’re early in that journey.”

Military-specific versions of ChatGPT and Grok recently joined Gemini on the platform that gives defense workers access to generative-AI tools for information that is sensitive but unclassified. DLA gets its workers up to speed on these tools through “GenAI 101 training,” Roberts said.

“I would look at generative AI as the concept of making the workforce and agency—from an operating model—comfortable with the results you’ll get with AI,” he said. Tools like Gemini and ChatGPT “allow the workforce to interact with AI, see the results, trust the results, and get comfortable.”

“That’s required, because as you move to a more agentic AI, those agents are working 24/7 behind the scenes, and so there has to be a level—a cultural level—of trust in those outcomes,” he said. “Now, we’re shifting to what I would call strategic implementation in terms of those agentic agents to get us better results.”

Roberts said introducing digital employees will require new management strategies. Humans will be working alongside—and often overseeing—digital agents, creating an assortment of new management challenges.

“I think we like to call it, ‘How are we managing our digital employees?” Roberts posited. “Because that’s what they really are, and how do you manage them in conjunction with your human employees?” How do you train a workforce to where everyone is essentially working as a supervisor because of the employment of AI with humans? Do you terminate an agent because of the performance?”

“So, it's a very complex universe, from an employee-cultural perspective, in terms of how you deploy and manage those agents,” Roberts said.

It might sound like science fiction, but the rapid evolution of AI—and DLA’s gradual adoption of it—creates questions around governance, rules, security, and the performance of AI agents and even human employees.

Many of these questions fall to DLA’s AI Center of Excellence, founded in 2024 to ensure the safe, secure, and responsible integration of AI across the agency. The center oversees DLA’s use of AI and performs other functions, like ensuring agents are being trained properly. DLA has implemented a zero-trust cybersecurity framework that Roberts called “foundational” to high-level agentic work.

Currently, the agency is establishing “persona-based access” that will dictate AI agents’ account permissions based on their job descriptions. For example, a procurement specialist at DLA may require access to certain data domains, like procurement and financial data, so a digital agent performing the same procurement specialist duties should have permission to access only those same data domains.

“These AI agents should have a persona, just like a person,” Roberts said.

DLA is also working to clean up its datasets.

“How do we create a single source of truth for DLA and our stakeholders in terms of our data, making sure it’s a gold standard, meaning we remove the duplicity, errors, etc…so the data is in a usable state, not just for analytics and AI, but as we look toward future system sustainability,” Roberts said.

‘AI is the new gunpowder’

DLA accelerated its use of AI during the 32-month tenure of Director Lt. Gen. Mark Simerly, who will retire on Oct. 2. Simerly will be replaced by a familiar face at DLA, Air Force Lt. Gen. David Sanford, who spent the past two years as director of Logistics Operations and commander, Joint Regional Combat Support.

Speaking at the Sept. 9 event, Simerly said technology allows DLA to be predictive and proactive in solving problems, and he views AI as the key ingredient to solving them at scale.

“I believe when it comes to military logistics, AI is the new gunpowder,” he told reporters. “And we don’t think it’s just an explosive, destructive capability, although we know it can be. It can also be very productive.”

Simerly’s philosophy at DLA has been “to move at pace with the commercial world,” adopting and adapting capabilities rapidly, establish the proper network and infrastructure to access and manage agency data suitably, “and then allow interoperability between the services and with industry to allow us to truly operate end-to-end supply chain operations that are optimized.”

A legacy of change

Simerly said changes at his agency were spurred by the war in Ukraine, which expanded in February 2022, and even during the seven-month-old war on Iran launched by President Donald Trump.

“I think probably the greatest change that I've seen in the agency is a willingness to change within the agency. This recognition that we have to respond to the changing environment, that we have to respond to the changing technology, and that we've got a responsibility to change as rapidly as we can, so we can continue to be relevant and offer support to the warfighters,” he said. “I would say that we were built to change from the very beginning. We've had to accelerate that change cycle in the last few years, and I've seen the willingness and the readiness of the agency to take that challenge on and to operationalize our transformation planning into tangible outcomes.”

The contested logistics in Ukraine captured his attention because they were so different from previous U.S. military conflicts in Iraq and Afghanistan, where DLA delivered food and fuel to forward operating bases “really without great resistance.”

“That would never take place in a place like Ukraine, in the way that they’re operating, the way they have to disperse, the way that the lines of communications are contested and where everything has to be done in a certain way,” Simerly said.

“What sort of sparked some of the transformation energy was very much awareness of what was taking place and being illustrated on that battlefield in Ukraine about the things that we no longer could rely upon, like the air and domain dominance that we possessed in previous fights that is now no longer valid or in place.”

Simerly said similar conditions pertain in the U.S. war on Iran.

“There's no question that we're learning the same lessons now in Operation Epic Fury with Iran and their ability to target a broader array of things with precision that puts much more at risk, and you know certainly we're learning a lot about the execution of alternate supply chains, alternate supply means, and the necessity of being able to operate with precision,” he said.

Simerly expressed pride in his DLA team.

“I think about throughout my career the winning teams that I've been privileged to be part of, and that's what I'm proud of here,” he said. “The winning team that I'm part of now, that will continue on the mission after I leave.”

]]>

AI-powered biowarfare is coming; the Army lays plans to ‘fight through’ it

Patrick Tucker — Wed, 09 Sep 2026 17:12:01 -0400

FORT BELVOIR, Va.— AI is “dramatically lowering the barrier” to creating biological weapons, according to the Army’s updated biodefense strategy—even as it gives commanders and troops a better picture of how biological threats, both natural and manmade, will shape future battlefields, service officials say.

The new strategy aims to prepare the Army to “deploy, fight, and win decisively in a bio-contested environment” by 2035, via new training, advanced early warning, and hardened equipment. But its leaders and bio-threat experts are already working to overhaul how the service thinks about and responds to biological threats. AI will be central to this new era of biodefense, whether advising in the disposition of forces or countering enemy attempts to sow confusion and panic about outbreaks, according to officials who spoke to reporters Tuesday.

The era of AI bioweapons

In 2023, Anthropic CEO Dario Amodei told U.S. lawmakers that AI could “widen the range of actors with the technical capability to conduct a large-scale biological attack,” a scenario he said could come to pass within “two to three years.”

The scenario is drawing nearer. This week, a new paper out of Stanford described how feeding genetic data into high-performance AI models could lead to a new kind of virus that can infect bacteria, showing it’s ability to breakthrough the immunity barriers of other organisms.

This has alarmed RAND molecular biologist Steph Guerra, who wrote, “As AI capabilities advance, viral bioweapons could become much easier to design and produce and could cause even more harm than the weapons used in the past.”

But AI also play a role in defending societies and militaries from such weapons. “We know that everything is changing faster now and so part of the key to the strategy is making sure that we're adaptable,” Alison Tamasi, the chief of the survivability and effects division at the Army’s Nuclear and CWMD Agency, said Tuesday.

Just as detecting a needle in a haystack is harder than making needles, using AI to detect biological threats on the battlefield is more complicated and human-centered task than feeding data into a model. That’s one reason why the updated Army biodefense strategy is so important, the researchers said.

One key element is new, portable genetic-sequencing kits. No bigger than a computer printer, the kits can analyze genetic samples on a laptop computer not connected to the cloud. That might sound like the sort of thing you could encounter in a doctor's office. But a cheek swab that a doctor might take in a relatively clean conditions is very different from samples contaminated with dust, air, fiber, or animal tissue. And getting analysis in the field in a day or so beats waiting weeks for a sample transported back to the States.

The Army hopes the kits, which can classify a sample as a known biothreat or a new one, will help forge strategies when new ones arise.

You can’t stop biology. You can plan for it.

AI isn’t the only factor increasing risks of future biological threats. New viruses occur naturally and spread more quickly today as a function of globalization. But the military’s response to the COVID-19 pandemic “exposed shortfalls in Army biodefense capabilities,” according to the new strategy.

A 2023 study found that COVID-19, which thrust the military into new civilian support roles and imposed new guidelines on forces around the world, undermined team cohesion, military readiness, and of course, mental and physical health.

The new strategy emphasizes better planning for operating during a pandemic, manmade or not. Demonstrating this capability might just deter an adversary from attempting to launch a bioattack as part of a larger campaign, officials said.

“The challenge is going to be how we do this [sample collection, analysis] in the crisis phase, or worse, in the conflict phase,” said Col. Dennis Ruffolo, deputy director for force health protection at the Office of the Surgeon General and Army Medical Command. “That's really what this is going to get after is is is making sure all Army leaders understand that we have to learn ways to fight through the biothreats.”

Here, too, AI will play a key role, said Ruffolo, in helping commanders better understand the biothreats they face, or even will face, in as much detail as possible.

That sort of understanding, enabled not just by new AI models but also humans collecting samples, monitoring environments, and performing analysis, will better prepare the Army for those contingencies. That adaptation means getting more soldiers to understand that bio-threat monitoring is a new priority, ensuring they are capable of biological threat monitoring, and providing them with the equipment and training to do so.

“I have operators below me who do not have doctorates or [are] not microbiologists, I have E-4 specialists who are able to operate that technology with protocols that I have written, and they can do that unsupervised.” said Army microbiologist Amina Zeidan, with the 44th Medical Brigade.

The strategy makes those sorts of efforts a core priority across training, equipping, facilities, policy, and so forth.

“Historically, the Army treated biodefense as a specialized problem,” reads the strategy. “Victory in future conflicts requires an Army that treats biodefense not as a secondary concern, but as a core warfighting imperative. This strategy provides the roadmap.”

]]>

Maven is becoming the Pentagon’s everything app

John Croxton — Wed, 09 Sep 2026 15:24:57 -0400

The Pentagon is finding ever more uses for the AI-powered command-and-control system called Maven Smart System, including logistics, force readiness, supply chain and budgeting data, the Defense Department’s AI chief said Tuesday.

Maven, which is used by all U.S. combatant commands for intelligence and operations data analysis, drew attention in the early days of President Donald Trump’s war with Iran by facilitating strikes on thousands of targets in lightning-quick succession.

But the war also revealed DOD’s need for a better way to plan operations against a 90-day picture of its munitions inventory, Cameron Stanley, chief digital and artificial intelligence officer, said at the Billington CyberSecurity Summit in Washington, D.C.

Already, MSS has replaced “six, eight, ten” different IT systems that U.S. military personnel previously used to analyze data, said Stanley, adding that the goal was to build a “golden thread” of data from sensors, to aggregation and visualization, to decision-making and acting based on those decisions.

While MSS started as a tool to bridge intelligence and operations efforts, the department is working to include readiness and disposition of forces, supply chain and budgetary considerations, modeling and simulation to improve decision-making processes—what Stanley called a “horizontal integration” of directorates’ data functions.

Stanley also pointed to GenAI.mil, the Pentagon’s main portal for AI tools, which recently added tools from OpenAI and xAI in addition to Google’s Gemini. It’s deploying “real, no-kidding frontier AI algorithms, and the full user experience, inside of a government environment,” he said.

The organizations that have been most successful are integrating AI in new ways and changing how they operate, versus simply automating existing processes, Stanley added. Whereas commanders used to wait hours for new satellite imagery, they now see images complete with target detection in minutes, requiring a rethink of military decision-making processes.

“Basically, we need to make better decisions faster with data,” Stanley said. “Everything seems confusing until you break it down to making better decisions faster.”

He also discussed procurement timelines, noting that recent use of other transaction authorities and marketplaces had shortened the time vendor selection took “from months to single-digit days,” compared to traditional requests for information and proposals at DOD. He recommended vendors use Tradewinds, DOD’s digital marketplace, saying it “satisfies the competitive requirement” with a five-minute video, “way better than writing a bunch of RFIs and RFPs.”

The department is also exploring AI pilot programs for drafting and concluding contracts, as well as legal advice, though humans would remain in the process. Stanley said that “90%” of the language used in other transaction agreements or Federal Acquisition Regulation-based contracts was standard, and AI could often be used to make straightforward updates as needed.

He noted that interim authority to test and authority to operate on department networks were “one of those clear bottlenecks that are really really hard to get around,” and that AI pilot programs were being trialed to speed approval times.

While the full timeline from a broad agency announcement to authority to operate right now “could be 18 months,” his office is “progressively looking at every single piece of that schedule.”

The department was historically “standoffish” about software delivery models like software as a service and platform as a service, but “meeting the industry where they are” would allow the Pentagon to invest in and integrate commercial offerings more readily, Stanley said. He said that DOD would adopt commercial AI models, use its own data to customize them for its own use cases, and build its own bespoke capabilities as necessary.

Instead of owning the software, “I'll pay a license for that, so you don't have to worry about development. We can just iterate inside of that license period,” he said. The department is “a commercial-first organization.”

Allies often “are aligned” but “don’t have the resources … the experience … the scale that we do,” Stanley said. DOD is actively working with partners including the Five Eyes (the U.S., U.K., Canada, Australia, and New Zealand) and “a few” NATO and Asian allies to understand their needs and help them avoid making mistakes DOD has already made, he noted.

Stanley added that “AI in everything” would happen naturally, regardless of DOD’s actions. The department would need to digitize older processes and integrate AI into all decision-making to adapt. “The way we've always done it is not the least risky option. The way we've always done it is the most risky course of action,” he said.

]]>

NSA wants AI to help analysts sift vast data troves

David DiMolfetta — Tue, 08 Sep 2026 22:20:00 -0400

The National Security Agency is leaning into artificial intelligence tools to process the mass volumes of data it collects, allowing analysts to sort through tasks and find anomalies faster, a top official said Tuesday.

“I think right now we’re looking at how to use an AI where we could triage so much more,” David Imbordino, the head of the agency’s cybersecurity directorate, said on a panel with officials from the “Five Eyes” intelligence alliance at the Billington Cybersecurity Summit in Washington, D.C.

The agency ingests vast amounts of data every day, including intercepted foreign communications and electronic signals from radar and weapons systems. Imbordino’s directorate, in particular, focuses on both intelligence collection and cyberdefense.

“Volume has always been a problem,” he said, but AI has raised the prospect of getting “to the nuggets” faster than before. In the near future, analysts will likely use AI tools to look at disparate pieces of information and evaluate them “much more quickly” than current methods that can take days or weeks, Imbordino added.

His remarks highlight one of many ways the nation’s cyber and intelligence agencies are putting AI to work and show the omnipresent role such systems will play in day-to-day work in spy offices.

NSA has also taken a leading role in the Trump administration’s AI policy efforts as increasingly powerful tools promise to help defenders find and fix software flaws while giving hackers new ways to break into networks. Late last month, NSA deputy director Tim Kosiba said that the agency wants access to a wide range of models and is engaging heavily with frontier AI companies.

The agency is helping to implement a sweeping June executive order directing national security agencies to create classified tests for determining when AI systems have developed sufficiently advanced hacking capabilities to warrant additional scrutiny. The order calls, in part, for a voluntary arrangement through which developers can give the federal government access to qualifying models for as many as 30 days before releasing them to other trusted partners.

]]>

CIA cyber intelligence helped US forces capture Maduro, deputy director says

David DiMolfetta — Tue, 08 Sep 2026 22:10:00 -0400

CIA cyber operatives provided intelligence that helped U.S. forces locate and capture deposed Venezuelan leader Nicolás Maduro within four minutes of landing on the ground in January this year, agency Deputy Director Michael Ellis said Tuesday.

Ellis credited the agency’s Center for Cyber Intelligence with building what he called a “flawless intelligence picture” for Operation Absolute Resolve, the U.S. operation that seized Maduro.

“If you look at [Absolute Resolve], you know that operation was only made possible by a flawless intelligence picture, and that flawless intelligence picture was built on cyber operations enabled by our CCI team,” Ellis said, speaking at the Billington Cybersecurity Conference in Washington, D.C. Without that intelligence, he said, the agency could not have helped U.S. forces locate and apprehend Maduro as quickly as they did.

The remarks offer a more detailed public picture of how CIA cyber prowess helped U.S. forces locate and capture Maduro, expanding on earlier accounts from officials that acknowledged support without explaining the agency’s entire contribution.

Joint Chiefs Chairman Gen. Dan Caine previously acknowledged Cyber Command’s participation in the efforts and credited the CIA, National Security Agency and National Geospatial-Intelligence Agency with supporting the Venezuela mission. He described months of intelligence collection on Maduro’s movements and daily routines, but did not directly identify the CIA cyber role Ellis outlined.

During the capture mission, multiple intelligence agencies established crisis action teams to support Special Operations Command and Southern Command. A U.S. official familiar with the operation previously said the NSA provided geolocation support and monitored communications and signals that could reveal troop movements or plans to activate radar.

Ellis discussed the operation in explaining the CIA’s recent decision to elevate CCI to a mission center, a change he said gave their cyber operators greater priority and helped align resources around intelligence needs.

The change has helped the agency “bring cyber operations to bear to solve these tough intelligence problems,” Ellis said. He did not specifically identify the systems targeted or explain how the cyber operations revealed Maduro’s location.

The CIA primarily executes and coordinates human intelligence gathering overseas, often done undercover. Officers recruit and manage foreign assets to clandestinely gather intelligence on areas like economics and terrorism threats. Much of that work has often involved the use of technology, including computer intrusions.

]]>

USAF speed up Reaper replacement after heavy losses in Iran

Thomas Novelly — Tue, 08 Sep 2026 17:15:30 -0400

The loss of dozens of MQ-9 Reapers to Iranian air defenses is speeding up the U.S. Air Force effort to replace its venerable strike-and-sensor remotely piloted aircraft with something cheaper yet better, service leaders say.

“As we get to the end of the MQ-9 life, there was some early talk of where do we want to go afterwards, and candidly, as a result of some of the losses that we've experienced in combat operations, that discussion sort of got accelerated,” said Lt. Gen. Christopher J. Niemi, the Air Force’s chief modernization officer, at a Thursday roundtable with reporters.

The service has lost at least 45 MQ-9s during Donald Trump’s war on Iran, roughly one-quarter of its fleet, the Washington Post reported in August. It’s not clear just how much that fleet cost, but an Air Force fact sheet puts the unit cost of one MQ-9 variant at $51 million in 2011 dollars—or nearly $82 million today.

Niemi said the service wants to buy at least 180 MQ-9 replacements, dubbed Massed Modular Aircraft, or MMA, and at a much cheaper price point. He said the price of one MMA should be about half of the $20 million Collaborative Combat Aircraft, whose first production contracts were awarded in June, four months ahead of schedule.

But those rough price goals—$20 million for a CCA, $10 million for an MMA—are for bare airframes, not the pricey sensors and systems that enable the drones to do their jobs.

As a service spokesperson put it later, the numbers “do not refer to total missionized costs.”

Niemi said this reflects lessons from past and ongoing wars.

“What we're seeing, in part based on our assessment of the threats around the world, and in part based on our real-world experience, is that you're going to lose more military hardware in future conflicts than you have historically, and that's what's driving us to this sort of fundamental principle that we need things that are easier to produce, cheaper to buy, and that we can afford to lose more of them,” he said. “Right now, we don't have a whole lot of that, so we're sort of trying to move the needle by bringing on things like MMA, so we have more options for the warfighter in the future.”

For that $10 million price tag, the service also wants an increased payload of 2,800 pounds and an unfueled combat radius of 2,300 nautical miles and 8,000 nautical miles for a one-way mission, according to a July Defense Innovation Unit solicitation. General Atomics, the Reaper’s manufacturer, has a souped-up version of the MQ-9, the SkyGuardian, which boasts a larger wingspan, range, and endurance. But top service officials don’t seem as likely to bet on it.

“One of the options that would be considered would be something like an MQ-9B, which is basically an updated and improved MQ-9A, but there are some reasons that we don't think that's the direction that we want to go as an Air Force,” Niemi said. “The main reason is, and and we're seeing this in our experience in the Middle East and as we look at threats around the world, that you need to have a portfolio of assets that's more attritable than the ones we've historically had.”

General Atomics spokesperson C. Mark Brinkley defended the Reaper’s service record and pointed out that the Air Force’s top uniformed leader has praised the aircraft’s use during Operation Epic Fury. He also said the company has worked with the Air Force “to hit cost targets for three decades” and that it would continue with other programs.

“The MQ-9’s unique combination of endurance, payload, range, sensors, weapons, connectivity, adaptability and reliability is precisely what enables mission effects that others struggle to replicate,” Brinkley said. “Today’s Reaper isn’t the same Reaper that rolled out 25 years ago. It’s been enhanced and upgraded many, many times. It’s also been called the MVP of Epic Fury, for good reason. We can debate many things, but at least we all agree that it’s a special aircraft that gets the job done.”

The Senate Armed Services Committee is backing the platform in its version of the 2027 National Defense Authorization Act by placing limits on how many can be retired and calls for more to be purchased by 2028.

General Atomics has also reportedly teased a new MMA offering with a 10,000-nautical-mile ferry range and claims that it can carry four Long Range Anti-Ship Missiles. Brinkely, while still defending the Reaper, also boosted its next-generation aircraft.

“Our new Wildfire aircraft addresses all of the concerns posed by the MMA solicitation. We can deliver years earlier and maintain the same high quality,” Brinkley said. “You can have everything that makes Reaper special, plus more, at a price that doesn’t break the bank. Reaper is special. Wildfire is special, too.”

]]>

The US military gets its own ChatGPT today

Patrick Tucker — Mon, 31 Aug 2026 09:00:00 -0400

Updated: 5:29 p.m.

The Pentagon’s main portal for AI tools now offers ChatGPT and Grok for Government alongside Google Gemini.

A Defense Department official said the OpenAI product, dubbed ChatGPT Mil, would bring new strengths to GenAI.mil, where troops and defense civilians can use generative-AI tools to work on controlled, unclassified information, or CUI.

“Gemini, I like for search functionality or scraping certain sources. Chat for more, like, text-based work,” said the official, who requested anonymity to speak before the official announcement on Monday. “We want to have access to all these different models, so people, if they prefer one for a certain task, they can use it.”

Work began in February to create a version of ChatGPT that could be used for CUI. In June, an OpenAI official said they believed it would be deployed in “early July.”

A DOD press release slated for release today said ChatGPT Mil would bring “a familiar commercial experience into the Department’s secure environment, tailored to warfighter needs...ChatGPT Mil supports document-heavy unclassified work across the Department, including planning, policy, logistics, and administration, and is built for scale to support more than 3 million Department personnel. This deployment accelerates routine tasks and frees up time for more critical projects across the Joint Force.”

One OpenAI official, who requested anonymity to speak before the official announcement, offered ideas for how the military might use its special version of ChatGPT. Operators at Northern Command might use it to streamline planning, resource allocation, or disaster response with FEMA or other parts of the federal government. Or the Joint Staff’s logistics directorate might use it to simplify paperwork and documentation. The official also suggested that logisticians at combatant commands “would be a natural audience, with uses such as reconciling supply and maintenance updates, drafting transportation requirements.”

In background talks, OpenAI and Pentagon officials said they don’t want to use models on the GenAI.mil platform for combat. The Pentagon has argued that they should be able to use AI products as they like, a sticking point in disagreements between the department and AI model maker Anthropic. And OpenAI has said that ChatGPT’s need for connectivity and large amounts of computing power doesn’t make it a good fit for combat anyway. The department is continually testing and seeking a number of other tools for those types of operations and actions.

Customized for security

Like the GenAI.mil version of Gemini, the military version of ChatGPT has additional security functions built in.

“When you're dealing with CUI within the department, it's obvious this is stuff that cannot be on a commercial model. We have to build it in a separate sort of environment. So it's engineered in a way to be separate from the commercial model,” said the defense official.

Prepping a frontier model for use on GenAI.mil is an “arduous” process, said one former defense official who was involved in such tests. It requires building out the “environment”—that is, finding the compute resources, keeping track of the users and their unique security conditions, compiling the data that the model will use, and anticipating the sorts of requests that users may have.

“Think of having to pass security guardrails for both DISA [Defense Information Systems Agency] and the NSA,” the former defense official said.

As DISA builds the environment, NSA looks for ways that attackers might invade it—or the model might escape it, the official said.

The Pentagon launched GenAI.mil on Dec. 9 with one tool: Google’s Gemini. Around the same time, Pentagon officials signed an agreement to bring Elon Musk’s xAI tool to the platform, but no formal deployment has since been announced. On Monday, the Pentagon announced that Grok for Government, an AI product of Musk's Starshield AI, was now available on GenAI.mil.

Joe Larson, who leads OpenAI’s government sales, said, “ChatGPT on GenAI.mil will make unclassified work throughout the Department more efficient, freeing up time and resources so DoD personnel can focus on the mission. This integration is grounded in our national security principles, pairing access to powerful tools with clear safeguards, meaningful human judgment, and accountability for high-stakes decisions. We’re excited to see what the Department’s extraordinary workforce builds with it.”

]]>

Can AI help the Marines simplify sustainment in the Pacific?

Jennifer Hlad — Fri, 28 Aug 2026 15:50:49 -0400

When a Marine planner is trying to determine whether a deployed unit has everything it needs, they must query several separate databases while cross-referencing supply levels, maintenance records, and more to create a spreadsheet. Marine Corps Forces, Pacific, is hoping an AI-powered logistics platform can make that process much faster and easier.

Defense tech company Tagup on Thursday announced a modification to its $2.1 million Small Business Innovation Research contract that will extend the use of its Manifest platform to certain theater sustainment cases in support of MARFORPAC. The platform is already being used by a handful of Marine units at the tactical and operational levels—and the company said its application in Marine medical logistics has resulted in a 25 percent reduction in purchasing costs and a 30 percent reduction in materiel handled, “with no compromise to readiness.”

Jon Garrity, CEO and co-founder of Tagup, said Manifest pulls from a variety of data sources to help Marines decide “what materiel, what resources do you posture where and when, and ultimately tying it back to measures of output or outcomes.”

On the supply side, it looks at the industrial base, he said. “We loaded over 4 million vendors in platform, we’ve got over 16 million parts represented. What are the lead times for each of those? How do we think about that? What are the locations of those vendors? What are the sources of supply?…and then the demand side is also, of course, critical.”

Historically, Garrity said, “at a theater level, a lot of this analysis was done at a headquarters level, not driven from the micro data. Imagine, for ground vehicles, you have every single maintenance action that’s ever been taken....That’s not used, generally speaking, to build theater-level [courses of action], and so that’s now the opportunities. Hey, we have technology and capability, AI capabilities that allow us to take the micro data, every single service request, every bit of inventory on the demand side, and then connect it to the supply side. What do we have? Where is it? Where’s the industrial base to be able to connect the two?”

Sustaining forces that are spread across the vast Indo-Pacific is a challenge, a MARFORPAC spokesperson told Defense One via email. “Marine units are highly dispersed, which intensifies standard logistics friction points.”

To overcome that challenge, the command “is testing how a unified logistics decision-support tool might automate” the data aggregation that could otherwise take several days of work, “to see if the software can ingest those disparate data streams and generate comparative courses of action—quantifying the impacts on readiness, cost, and operational risk—in near real-time or a matter of hours.”

Garrity said Manifest can also help Marines better use the materiel they send forward, “so as a result, you’re maintaining higher levels of readiness for the end items that that materiel supports, you’re… making better use of the space available on a [deployed ship], and you’re typing up less capital, less cash, in the materiel that’s not being used effectively.”

The platform, which the company calls “an AI-powered multidimensional logistics decision engine” has a conversational AI interface called Argus that allows users to ask questions “and explore scenarios in plain language.”

]]>

Privacy-focused communication app aims for military customers and beyond

Patrick Tucker — Thu, 27 Aug 2026 22:39:29 -0400

HONOLULU—A new encryption-based app, Kibu, available to U.S. military operators and foreign military partners, could change the way militaries and consumers communicate.

Civilian and government officials are accustomed to using their own phones and tend to rely on commercial messaging apps, even when they shouldn’t. For example, U.S. operators in Poland and elsewhere have communicated with Ukrainian counterparts over the popular communication app Signal, because of its wide adoption and availability on regular consumer phones.

But those conversations—particularly those involving U.S. special operators—may occur over networks that the U.S. military doesn’t own, so-called dirty networks. This is because they often conduct operations with allied militaries that don’t have access to U.S. top-secret communications platforms or equipment, a challenge that participants at a recent irregular warfare symposium here highlighted across multiple discussions.

“Government executives, they tend to be using their personal devices. They're on Signal,” Eftychis Gregos Mourginakis, co-founder of secure messaging app Kibu, told Defense One during the Global SOF Indo-Pacific Irregular Warfare Symposium.

But there’s no guarantee that the individual on the other end of the phone is the right person, or even a person at all, he said. “How [do State Department officials] trust that when they're talking to their British compatriot, that it’s actually that person? Like, how do you fill that gap? You can't do it with voice. You can't do it on video.”

Recent incidents involving AI agents posing as humans have highlighted the growing trend—one that a current military official who tracks adversary disinformation campaigns said is a big problem, especially for the military and intelligence communities.

“AI has made deception scalable. We’re no longer just fighting phishing emails, we're fighting synthetic voices, cloned faces, and entire fake identities convincing enough to fool the people who know the real person best,” the official said.

A small number of U.S. special operators are now testing Kibu, which uses encryption and other verification features to make sure that person—or AI agent—you’re talking to is who they say they are. That includes not just device-to-device encryption, but also biometric authentication and, for military members, can include other forms of identification as well. Organizations can add their own, such as the CAC card requirement. The app even has special passcodes in case the user has their own phone or device but is in a situation where someone else is trying to access it against their will.

The app isn’t tied to a device SIM card, phone number, or even PIN number. And different organizations can host it in different cloud environments. “So if an organization basically signs an agreement with us, all of their end users on this platform are totally air-gapped,” Mourginakis said.

And even as the military faces the challenge of keeping adversary AI agents out of sensitive conversations, they also want to use more AI under what the military calls “meaningful human control.”

Mourginakis said the app allows individuals to build their own agents within the environment to communicate with unmanned weapons but still report back to the human operator.

“The identity can look like a human identity, but it can be sandboxed by the human users in that space. They need to biometrically prove it,” he said. “You can see we have a decision quorum. So if I need in this group to authorize certain things, say a certain number of these people need to approve these actions, here they can communicate about it, they can talk about it, they can share the intelligence they need, and then go, hey, yeah, I'm good to do this. Commit my decision. Thumbs up. It's approved.”

While military members face the most immediate need for that sort of verification, Mourginakis said, secure communication is a growing priority for civilians as well.

He sees a shift away from tech products that monetize consumers’ data and push sharing of content toward companies that prioritize privacy.

“Now because of that and the rise of AI slop, the biggest premium service is going to be privacy, trust, security as a service.”

]]>

US military bought commercial satellite images of alleged narco airstrips, Chinese naval fleet

Thomas Novelly — Thu, 27 Aug 2026 16:54:02 -0400

The U.S. military purchased satellite images that identified Chinese aircraft carrier positions and Ecuadorian airstrips allegedly used for drug trafficking, a new report from the Government Accountability Office revealed. But despite those and other examples of prior use, the Space Force still hasn’t fully embraced commercial intelligence, the watchdog said.

“Some Space Force officials told GAO about challenges they have faced related to purchasing and using commercial space data such as licensing costs, perceived use restrictions, and concerns over long-term access to the data,” the report said. “These challenges led to hesitation to purchase and use commercial data among these potential users.”

The Space Force’s Joint Commercial Operations Cell, the group that buys the majority of the service’s data, spent nearly $77 million between 2023 and 2025 on commercial space services through the Global Data Marketplace, an e-commerce site used by the military for unclassified images. The JCO has a working group to discuss purchases being made by the U.S. government, but Space Force officials aren’t included in that group, according to the report. That’s led to the hesitations and confusion about purchasing intelligence on the platform, the GAO said.

“This working group meets monthly and coordinates purchases. GAO found that many of the challenges reported by Space Force officials could have been addressed with discussions at the working group, but the Space Force officials were not part of the working group and information about the working group is not widely available,” the watchdog report said.

The GAO recommended that Air Force Secretary Troy Meink, who formerly served as principal deputy director of the National Reconnaissance Office, increase communication and awareness of the working group. The Defense Department concurred with the watchdog’s findings in a June 18 response.

“Increased communication about the working group could lead to additional Space Force and other officials engaging in discussions that could address perceived issues and improve the use of commercial space data and services for national security purposes,” the GAO report said.

The Space Force will need to rely on those commercial assets for future fights, according to the GAO. As adversary counterspace capabilities grow, the Defense Department needs private satellite imagery to fill in blind spots and decrease risk to its own space assets.

“Threats to DoD’s space assets from strategic competitors have grown substantially in number and intensity,” the GAO report said. “DoD is working to integrate commercial space solutions, such as data and services, to mitigate these threats as well as enhance its capabilities.”

Jon Ludwigson, GAO’s contracting and national security acquisitions director, said in a phone interview on Thursday afternoon that there are positives to using those commercial systems in tandem with the military’s highly-classified spy satellites.

“We're not going to use commercial data for everything, but if we can use it for some things, it reduces the pressure on the exquisite systems to do things,” Ludwigson said. “It also gives us the opportunity to cue others that we're watching in ways that don't require us to disclose that we're using exquisite systems.”

The National Reconnaissance Office, an intelligence agency which builds and launches spy satellites, purchased commercial electro-optical imagery from Vantor, Planet Labs, and BlackSky through its $4 billion Electro-Optical Commercial Layer contract awarded in 2022, according to GAO.

Images purchased by NRO in 2024 showed “changes in Chinese aircraft carriers and other ships’ locations over time, and identify the types of ships” at Sanya Naval Base in Hainan province.

That same year, JCO purchased images that “helped the U.S. Southern Command identify possible clandestine airstrips in Ecuador.” The U.S. government shared the intelligence, which led the Ecuadorian government “to further investigate and disable the airstrips that were likely used by narcotraffickers,” GAO reported.

Identifying those sites and sharing their locations with Ecuador is an example of how governments can benefit from the unclassified imagery, Ludwigson said. It also protects and frees up use of the military’s most crucial spy satellites.

“When we go to share stuff that's coming off an exquisite system, that image is, itself, classified,” he said. “So, if we can use commercial imagery for those conversations with the Ecuadorian government or other governments to say, ‘Hey, look at this. This is what we think it is, and this is what we are concerned about,’ it’'s a commercial asset, then we haven't given up anything about our exquisite assets.”

]]>

NSA wants access to ‘all’ AI models, top official says

David DiMolfetta — Thu, 27 Aug 2026 12:33:00 -0400

The National Security Agency wants access to artificial intelligence models across the commercial market and is holding extensive discussions with leading developers as it moves to carry out new White House directives on the technology, the agency’s deputy director said Thursday.

“We want access to all the models, and we’re going to take advantage of that,” Tim Kosiba said on a panel at an Intelligence and National Security event in Bethesda, Maryland.

Kosiba said NSA has used AI in various forms for decades, but the rapidly improving capabilities of newer systems are driving a more aggressive effort to obtain and deploy the fast-evolving technology.

“The transformative change that we see today is what these models can actually do,” he said.

The remarks offer one of the clearest public indications that NSA is actively engaging major AI developers as it assumes a central role in the government’s new framework for evaluating advanced models. They also follow months of internal deliberations among officials regarding the role the U.S. intelligence community’s primary signals intelligence and cyber organization should play in advancing U.S. AI competitiveness.

“There are a lot of conversations happening amongst industry, especially the frontier model companies,” Kosiba said. He didn’t identify the companies involved, say which models NSA currently uses or disclose whether any developer has provided the agency with access to an unreleased system.

The term “access” may cover several different arrangements. NSA could use a model through a company-controlled service, deploy a version inside a government environment or receive access for security testing of classified networks.

The conversations are occurring as NSA helps implement a sweeping June executive order directing national security agencies to create classified tests for determining when AI systems have developed sufficiently advanced hacking capabilities to warrant additional scrutiny.

The order calls for a voluntary arrangement through which developers can give the federal government access to qualifying models for as many as 30 days before releasing them to other trusted partners. It explicitly prohibits the creation of a mandatory government licensing or approval process for new AI models.

NSA’s director is responsible for deciding which systems meet the threshold for a “covered frontier model,” in consultation with the Office of the National Cyber Director, the Cybersecurity and Infrastructure Security Agency and other government offices.

A related national security directive issued in June instructed intelligence and defense agencies to form “deep, proactive partnerships” with industry and make the most advanced models available to national security personnel without delay. It also encouraged the government to avoid becoming dependent on a single company by obtaining technology from a diverse group of suppliers.

Kosiba’s comments didn’t establish that the voluntary pre-release program is fully operational or that companies have begun providing models through it. The White House told major developers earlier this month that open-weight models would not be included in the testing program. Unlike closed systems controlled by companies like OpenAI, Anthropic and Google, open-weight models make their underlying settings available for others to download and modify.

Government access to advanced AI has expanded. OpenAI’s latest GPT-5.6 models became available to federal customers this month through its FedRAMP-authorized enterprise service after their public release in July.

The push for broad access comes as the Defense Department remains locked in a legal and policy dispute with Anthropic over restrictions the company placed on certain military uses of its models. The Pentagon designated Anthropic a supply-chain risk after the company declined to relax some of those safeguards, prompting a lawsuit in which a federal judge temporarily blocked parts of the government’s action.

The NSA has gotten carveouts with the company, and is already using versions of Anthropic’s advanced Mythos model on an experimental basis, according to reporting from The New York Times earlier this month. Current and former officials told the newspaper that the agency is using the technology to test defenses protecting U.S. military networks and evaluate its ability to identify weaknesses in networks belonging to China, Russia, North Korea and Iran.

Parts of NSA temporarily lost access to Anthropic’s Mythos 5 amid an export control dispute between the company and the Trump administration, demonstrating how government access to commercially controlled systems can be affected by policy decisions outside the agency. The administration has subsequently lifted the restrictions that had affected access.

AI-enabled cyber risks have become an increasing flashpoint in recent months after Anthropic unveiled its first version of Mythos, an advanced model it withheld from public release because of concerns that its ability to find and exploit previously unknown software vulnerabilities could be misused.

Since then, Anthropic, OpenAI and other developers have introduced a growing family of models with similar capabilities, forcing policymakers to consider how to test the systems, control access to them and harness their abilities for cyberdefense without giving attackers the same advantage.

As NSA pushes to adopt AI faster, Kosiba said humans will remain responsible for checking that its use complies with surveillance laws and the agency’s internal rules.

“What the models are putting out, how we’re leveraging technology today, will always require that human to double-, triple-check and make sure we are in complete compliance with the law,” he said.

Editor’s Note: This story was updated to add additional context from New York Times reporting.

]]>

What Trump’s tech strategy means for the military, deterring China, and the future of AI

Patrick Tucker — Wed, 19 Aug 2026 03:44:46 -0400

The newest White House tech strategy pushes faster development of emerging technologies for the military, gives defense innovators more room to experiment, and will broaden the federal market for young defense tech startups. But it also doubles down on a particular approach to developing AI—one that favors a handful of well-connected U.S. tech companies at the expense of other innovative approaches. That could benefit China and slow the U.S. military from getting the tools commanders actually want.

Here’s a breakdown of the winners and losers.

Winners: swarms, deployed AI, faster military technology

The strategy puts a big emphasis on new, tech-driven approaches to warfare, particularly the use of drones and even drone swarms alongside manned and legacy assets. Specifically, it calls for “optimal combinations of lower-cost (and sometimes lower-tech or attritable) platforms that can be deployed in larger numbers and complement more limited numbers of sophisticated platforms.”

Over the last decade, efforts to field autonomous aircraft alongside fighter jets or ships, for instance, have been uneven at best—even when the Pentagon changed its talking points on the necessity for large numbers of low-cost drones—constrained by programs' internal buying policies and previous commitments to other programs' investments. The Replicator program provides a vivid example, hailed as an essential pathfinder but receiving only $1 billion in funding.

The new strategy sets specific “priority areas” for new military research and spending: undersea, space, and AI and autonomy. These priorities are the most relevant to “deterrence in the Indo-Pacific”—as in, preventing China from launching a major conflict. Within those priority areas, it lists “multi-agent systems and swarm intelligence” and various uses of autonomy, from robotics to command and control, as “critical.”

The strategy also takes square aim at policies and practices that slow down military buying of new technologies. Building off other executive orders, it pushes Pentagon buyers to bypass traditional acquisitions in favor of a “mix” of contract types, such as performance-based contracts and other transaction authorities, or OTAs. It also resets the tone for military purchasing toward “faster non-traditional approaches that may involve higher risk but offer high potential reward,” something that commanders, younger defense tech leaders, and even lawmakers have long pushed for.

Perhaps most importantly, it also takes steps to actually grow the market for younger defense tech players beyond just the military. It takes aim at foreign military sales rules and export controls that hobble the sale of much defense tech to allies. That will be especially important for newer defense companies trying to secure early-stage investment to survive. And it calls for other federal agencies to build “streamlined pathways” for smaller companies to at least get their foot in the door, even via partnerships that don’t offer young companies much more than access to federal infrastructure, such as “Cooperative Research and Development Agreements (CRADAs) and Agreements for Commercializing Technology (ACTs).”

Losers: open weights, small AI

While the strategy champions smaller, more innovative players in defense, it also weighs in on the debate about what the future of AI should look like and comes down on the side of Big Tech companies that are pushing a specific, energy-intensive, and deeply unpopular future of AI over emerging strategies favored increasingly by researchers. China will benefit from the oversight, Michael Schiffer, a partner at Scalare Advisors and former Deputy Assistant Secretary of Defense for East Asia, wrote on Tuesday for Just Security.

In its section on critical technologies for support, the strategy lists “Foundation models, including large language, multimodal, and world systems.” The section also includes niche tech areas like brain-computer interfaces but omits entirely open-weight models and open-source software development.

Here’s why that’s important:

Foundation models, such as today’s large language models from companies like OpenAI and Anthropic, represent just one potential method for building AI tools—one that relies heavily on pushing as much data through as much computational infrastructure as possible to achieve something that looks like “reasoning,” but that’s really just plain old probability calculation.

Anthropic co-founder Dario Amodei helped to transform that concept from a research effort into a massive endeavor to acquire computational resources and data as quickly as possible. Amodei explained it simply in a 2023 discussion with Dwarkesh Patel. “It turns out that raw scale—more compute and more data—drives capabilities far more powerfully than complex algorithmic inventions or hand-crafted architectural changes.”

A small handful of frontier AI labs, and their big cloud computer backers, are in a literal race for company survival because there’s only so much power, so many chips to continue to pursue “raw scale.” If, as Amodei says, just having more “more compute” and “more data” means having the best performing models, then whoever has the most of those resources can push out any competitor.

It’s no wonder the AI race has become a concentration of wealth and power, in just a handful of labs with their large cloud compute backers. No wonder, also, that public perception of AI has plummeted: just 18 percent of Americans believe it will be a positive force for the United States, one new poll found. But the foundation model approach does have its fans: large-scale U.S. cloud providers who have invested billions in foundation model labs.

The very way the strategy frames the issue plays to the idea that AI has one future and that winning the AI race or achieving dominance means favoring a handful of players. But it’s also a theme that plays out in many of the White House’s policies and executive orders around AI.

Meanwhile, alternative approaches to building AI are emerging. Rather than race to build more data centers and find more data to throw at the problem of AI, they use parameters that already exist that are freely available, called open weight models, as opposed to closed weights that are protected intellectual property. And they use those in conjunction with alternative architectures that make using those models more energy efficient. Many of these alternative approaches to building AI perform almost as well as what Anthropic and OpenAI are offering, and have the potential to do so at far lower energy costs.

Why it matters to the military

The finite amount of compute power, data, and good will when it comes to AI could become a strategic problem for the U.S. military. If public sentiment undermines congressional and financial support for research and development in the field, officials fear, the country might lose its AI edge to China, which has openly declared its intention to pull ahead.

In the meantime, frontier-AI models also create tactical and operational problems for the militaries that use them. Troops and military units in the field can’t count on the connectivity that tools based on large models need. And in a new report for the Carnegie Endowment for International Peace, Jake Steckler writes, “Ukraine’s computational architecture, consisting of Western cloud access, domestic data centers, and forward-deployed compute nodes, is already straining as it integrates more AI into targeting and coordination.”

Open-weight models don’t appear in the new strategy, though it does list niche and far-fetched concepts like brain-computer interfaces as “critical.” That’s a huge oversight and a potential gift to China, argues Schiffer.

“China’s advances in open-weight AI have exposed the limits of a U.S. strategy built too heavily on the idea of denying them access to American technology and American markets,” he said.

]]>

White House S&T strategy calls for new approaches to safeguard US research

Edward Graham — Mon, 17 Aug 2026 18:20:00 -0400

The Trump administration says the nation’s science and technology research ecosystem needs to embrace “novel security solutions” to ensure the U.S. maintains its global edge, according to a new document publicly released on Monday.

The National Security Science and Technology Strategy published by the White House Office of Science and Technology Policy outlines how the U.S. innovation sector should be leveraged and guarded to defend the country. The new document builds off last year’s National Security Strategy, which called, in part, for “protecting the competitiveness of the U.S. economy and bolstering the resilience of the American technology sector.”

To help safeguard the nation’s innovation sector from threats “that could impair U.S. economic and military strength and competitiveness,” the strategy calls for “approaches that protect without inhibiting the productivity and agility of the national security S&T enterprise.”

President Donald Trump previously moved to curtail illicit foreign interference in the nation’s research institutions, most notably through a January 2021 memo that sought to keep adversarial governments from exploiting U.S. science and technology research.

Lawmakers from both parties have also pushed legislation to limit Chinese involvement in federally-funded research projects, and the Pentagon similarly announced on Monday that it called for 30 U.S. academic institutions to review their ties with foreign entities.

But the strategy says research institutions need to take further steps to enhance their threat awareness, including conducting “automated research security vetting” of agency-funded research proposals and performing continuous monitoring of federally-supported projects.

It also calls for “bolstering counter-intelligence support to governmental and non-governmental research centers” — which would give researchers more immediate and timely insights into potential threats—and creating “risk-based review criteria and a repository of sharable information to enhance the efficiency of information collection and sharing among agencies.”

It noted, for instance, that the FBI-led Quantum Information Science and Technology Counterintelligence Protection Team “brings an interagency team of quantum experts and security professionals to fine-tune technology protection and security outreach activities for the quickly evolving quantum R&D community.”

Another section calls for “establishing cybersecurity guidelines for researchers and research institutions,” although it is sparse on the details of what this would look like. The National Science Foundation, for its part, has a research security policy framework that includes some cyber-specific guidance as part of its broader blueprint.

“Given the evolving nature of research security threats, security experts should embrace comprehensive, new, and novel security solutions, particularly where they can be crafted to address unique or nuanced requirements of specific technology fields,” the document said.

The strategy’s implementation section highlights the need to enhance K-12 and higher education training opportunities and bring more skilled talent into the federal workforce—an effort the Trump administration has been undertaking, in part, by prioritizing skills-based hiring over degree requirements.

Notably, however, the strategy references the need to attract more nonimmigrant tech workers to the U.S., which differs from the Trump administration’s prior stance that called for a more restrictive approach.

“The United States will further strengthen its workforce by attracting and retaining top-tier global talent in critical national security S&T fields,” the strategy says. “Agencies will collaborate on using existing authorities to recruit the exquisite talent necessary to ensure the national security S&T Enterprise leads the world in developing and applying these technologies.”

OSTP did not immediately respond to a request for comment about the strategy’s inclusion of the need to attract and retain global talent.

Last year’s National Security Strategy said, in part, that the U.S. “cannot allow meritocracy to be used as a justification to open America’s labor market to the world in the name of finding ‘global talent’ that undercuts American workers.”

Trump issued a proclamation in September 2025 that moved to restrict the number of H-1B visas, saying that the program has harmed American workers. The visas allow U.S. firms to temporarily hire foreign workers, with the tech industry notably using the program to attract nonimmigrant talent. The proclamation attempted to institute a $100,000 H-1B visa fee, although a district court has overturned that mandate.

In addition to calling for the nation’s innovation sector to embrace new security solutions, the strategy said the U.S. needs to speed up its pace of innovation, build up its domestic “technological resilience” and focus its science and technology ecosystem on “maintaining battlefield advantage, countering strategic threats to the homeland, and shaping the competition toward areas of U.S. strength.”

Nextgov/FCW Cyber Reporter David DiMolfetta contributed to this story.

]]>

DARPA funds advancements in quantum networking

Alexandra Kelley — Thu, 13 Aug 2026 09:00:00 -0400

The Defense Advanced Research Projects Agency is exploring how to scale the reliable transmission of quantum data and pave the way for a viable quantum internet by giving networking networking company Qunnect an award to further deploy networks designed to handle qubit traffic.

Pursuant to the contract, Qunnect is granted new funding to further deploy its Carina quantum networking solution, the company announced on Thursday.

Qubits are naturally more fragile than classical bits, and transmitting them from one node to another has been a persistent problem in scaling a viable quantum network. Carina addresses several outstanding problems in safely transmitting quantum data by operating at room temperature, stabilizing environmental noise, maintaining entanglement during transmission and other engineering challenges.

Qunnect will specifically focus on refining Carina’s polarization compensation nodule, its internal system that helps reduce distortions in data transmission.

“Quantum networking has mostly stayed in the lab for decades because entanglement can be lost as a result of environmental factors like heat, wind and other everyday fiber movement,” Noel Goddard, CEO of Qunnect, told Nextgov/FCW. “Carina is the only commercial product built to cancel that noise out in real time, like noise-cancelling headphones for both buried and aerial fiber. This is one of the keys to Carina distributing entanglement over existing infrastructure instead of requiring purpose-built fiber or being limited to pristine lab environments.”

DARPA declined to comment on the award. Qunnect declined to disclose the exact amount of funding in the contract.

“We made an early commitment to designing instruments that operate on the same infrastructure the world already uses,” Goddard said. “For the past five years, we focused on making quantum networking practical, and we’ve validated it through deployments in cities, and through partnerships with private industry and growing support from government agencies. We know that reliability is what will separate proof-of-concept demonstrations from useful networks that enable tomorrow’s applications.”

Quantum networking has received more attention at the federal level recently as the quantum computing industry looks to reach new levels of maturity and successfully share data between fault-tolerant machines. In 2025, DARPA unveiled its QuANET initiative, which aims to marry components of classical and quantum systems together in a functioning hybrid network.

Goddard said that Qunnect’s award is a standalone Small Business Innovation Research award.

President Donald Trump released an executive order earlier this year focused on advancing quantum information sciences and technology research that specifically focuses on spurring the development of a distributed quantum computing network, tasking agencies including the Department of Energy, Department of Commerce, the National Science Foundation and NASAn to craft and submit five-year roadmaps documenting their efforts to expand quantum networking.

]]>

The lasers are coming, but don’t forget training

Lauren C. Williams — Tue, 11 Aug 2026 21:12:47 -0400

HUNTSVILLE, Ala.—Lasers and high-powered microwaves will need equal standing to traditional weapons when it comes to the Army’s approach to missile defense, the service’s space and missile-defense chief said. But as battlefield tech is developed and integrated, proper training will be key to keeping soldiers safe.

“Integrating systems like lasers or high-powered microwaves, there's a pretty significant burden of implementing that across training and organizations,” Lt. Gen. John Rafferty, who leads Army Space and Missile Defense Command, told reporters Tuesday at the Space and Missile Defense Symposium.

Rafferty’s top enlisted advisor concurred.

“I know everyone's talking about faster, cheaper—all those things are needed. But what's most important is our soldiers need to be trained on it,” Command Sgt. Maj. Rickey Jackson told reporters Tuesday. “We don't want to assume too much risk in that, because the soldiers are the ones that are going to have to use it and implement it. So I look forward to the rest of this week and hearing from industry and how we can do that, because ultimately our soldiers need it.”

In his keynote, Rafferty said air-defense strategy should balance the use of missiles, which can reach beyond line of sight, and energy weapons, which can be cheaper per round and used for closer ranges.

“We need a more comprehensive missile defeat strategy that focuses on interceptors—equally so on non-kinetic effects, both electronic warfare and directed energy—and the type of targeting [command and control],” on top of the ability to move sensitive data to lower classification levels “and make decisions to engage targets faster,” he said.

After decades of touting directed energy as a weapon of the future, the Pentagon may at last be on the verge of realizing its vision. Nascent programs such as the Army-Navy Joint Laser Weapon System and the Army’s Enduring High Energy Laser, as well as operational systems such as the Navy’s HELIOS drone-dazzler, suggest that directed-energy weapons may soon go mainstream.

The Army tested non-kinetic weapons during a recent test at White Sands. They synced C2 systems to get a clear picture of the battlefield and help soldiers pick the right weapon for the incoming threat, Rafferty said.

“The individual systems worked, right, because they're all systems that exist right now. That wasn't the trick. The trick was integrating them into a command-and-control system to make sense of and then assign effectors against those against those targets. So that, in a nutshell, is a sort of combined-arms approach to missile defeat or or a comprehensive missile-defeat strategy,” Rafferty told reporters.

“I'm not surprised that we could do it, I'm just thrilled that we actually did it, that we decided to integrate multiple C2 systems and figure out how to enable our soldiers to defeat complex threats.”

The Army has leaned forward on adding lasers to its armory through a directed energy program.

But ultimately, the concern isn’t about the Army’s ability to integrate new tech and weapons systems into its operations—it’s training.

“Clearly, there are technical challenges to integrating platforms into command-and-control systems and planning software that helps us to integrate the effects at the right place and time,” Rafferty said. “But in terms of our Army being well-equipped and agile enough to do this, I'm 100-percent confident.”

The challenge comes in making sure there’s an adequate power source, in “maneuvering and supporting” the system, he said.

“It's not about the laser anymore. It’s really about putting it together, packaging it, making it ruggedized, and coming up with the doctrine and the techniques…the sustainment that needs to be done to go from a high-end technological capability to a no-kidding weapons system that’s able to be…usable by 20-year-old sergeants,” Tom Karako, a missile defense expert at the Center for Strategic and International Studies, told Fox News in a recent interview.

That also means the systems have to work in bad weather and austere conditions, and be repairable in the field, he said.

]]>

Rough waters help firm demo ‘expeditionary manufacturing’ aboard warship

Jennifer Hlad — Fri, 07 Aug 2026 02:03:37 -0400

HONOLULU—Heavy seas didn’t stop Firestorm Labs from printing “thousands” of unique parts aboard the USS Essex amphibious assault ship on the way to the Rim of the Pacific naval exercise here, the expeditionary manufacturing company announced this week.

“We were told it was the roughest water off the California coast in the last five years,” Firestorm CEO Dan Magy told Defense One. The company has an exclusive arrangement with HP, which uses a type of 3D printing called multi-jet fusion, “and what we’ve found is that the systems and the machines are quite resilient.”

“We think we’re building a contested logistics solution that is going to be scalable across all branches of the Department of War,” Magy said. “If we’ve seen anything from Ukraine and now from Iran… we need to rethink warfare from the ground up. We are increasingly dealing with contested environments where—you know, look at what just a blockade in one part of the world has caused to the ability to get products out, oil prices. Heaven forbid something like that happened in the South China Sea, where 70 percent of the things we all buy in the world are produced.”

Magy believes their xCell additive manufacturing unit—composed of two 20-foot containers that can go anywhere—"really represents a leap forward in how we're thinking about solving, maintaining legacy systems, but also building attritable systems like drones closer to the need.”

During a tour of the xCell unit used aboard the Essex, Duane Blank, field operations manager for Firestorm, showed off some of the parts printed at sea, including stops made to protect Apache helicopter rotors from scraping against the ship’s non-skid coating.

The nylon composite the company uses for printing works for “everything within the entire spectrum of warfare,” Blank said, from prosthetics to drones to humanitarian assistance and disaster relief items.

Among the items printed while pitching through Sea State 5 were a dozen Squall first-person-view drones, Blank said, which were later used as adversary aircraft during a counter-drone exercise as part of RIMPAC.

The “beauty of 3D printing,” he said, is the “ability to iterate” and get troops what they need “at the point of need.”

Magy said it’s a top concern.

“The No. 1 thing we're hearing from folks across the service is, ‘It's great if you make it in a factory in Alabama or California or Texas. What happens if the thing breaks on the battlefield? I have to put it in a box now for four months.’”

Sustainment wasn’t what the company initially set out to address, he said, “but it has become increasingly important for all of our customers, because there are a number of legacy systems that like, the supply chain just isn’t there anymore. Let alone the supply chain for something in Okinawa or in Bahrain.”

]]>

Can an Alaskan spaceport ease Pentagon’s launch bottleneck?

Thomas Novelly — Wed, 05 Aug 2026 17:30:13 -0400

Rocket Lab’s announcement that it would launch suborbital rockets from Alaska could ultimately give the Pentagon a much-needed new option for national-security launches, the company and space experts told Defense One.

Last week, the Space Force’s acquisitions arm announced a $266 million, multi-launch contract with the California-based rocket company. That will cover 12 to 18 launches, which will start this year and “primarily take place from a new Rocket Lab location at the Pacific Spaceport Complex-Alaska (PSCA) in Kodiak," the company said.

These missile-defense missions will be handled by the company’s HASTE rocket, a suborbital variant of its Electron.

Both rockets “use the same launch infrastructure so yes, operating out of Alaska naturally opens the door to a broader range of possibilities including orbital launch,” a Rocket Lab spokesperson said. “It’s important to remember too that the core systems, infrastructure, and expertise we build and refine in Alaska aren’t tied to just one location.”

Top Space Force officials and lawmakers are concerned that the Cape Canaveral and Vandenberg bases will be unable to handle the projected increase in military-satellite launches. The Space Force expects to launch more than 100 rockets a year within five years and potentially thousands a year by 2036.

An Arctic national-security launch site would reduce the bottleneck and help the military reach polar orbits, said Clayton Swope, the deputy director of CSIS’ Aerospace Security Project.

“Putting something in Alaska, you kind of get both of those things,” Swope said. “You get this diversification away from the two main launch sites we have, and you are creating a launch site that is more optimized around a different type of use, which is related to the orbit, the inclination.”

The state-owned spaceport where Rocket Lab will host its new launch location boasts “a low population density, manageable air and maritime routes” that would distinguish it from some other launch facilities. But it has hosted just nine orbital missions since it was founded in 1998.

Charles Galbreth, a retired Space Force colonel who directs space studies at the Mitchell Institute, said the Rocket Lab expansion “could potentially be very helpful” for small and medium national-security launches—eventually.

“It's a capacity or a capability waiting for a demand signal,” Galbreth said. “So there's got to be the combination.”

Alaska can be an important addition, yet its isolation will present challenges, said Brian Weeden, who directs civil and commercial policy for the Aerospace Corporation’s Center for Space Policy and Strategy.

“I certainly think there is room for an expansion of different kinds of launch activity in Alaska. I don't think it's going to probably take the place of also standing in other areas,” Weeden said. “You're still going to need power. You're going to need to get rocket fuel up there. You're going to get the payloads transported up there, and that may be a little bit harder to do in a more remote location like Kodiak, as opposed to a place like Cape Canaveral or Vandenberg, where a lot of infrastructure already exists.”

]]>

As AI models break free, White House works with firms on secret safety measures

Patrick Tucker — Wed, 05 Aug 2026 13:25:04 -0400

No one is saying, officially, what was decided when top frontier AI labs—Google, OpenAI, Anthropic, and Meta—met with White House officials on Tuesday to discuss voluntary guidelines for testing new models.

But Democratic lawmakers described the Trump administration's approach to regulation as “ad-hoc and unpredictable,” and said it’s likely to boost global adoption of rival Chinese models at the worst possible time.

Two officials with one of the labs said that Google, Anthropic, and OpenAI submitted a joint draft of the regulation around nine days ago and then began to work with each other and with the White House to find points of agreement. According to the officials, the labs all agreed they should be able to continue A/B testing as part of the process for developing models, and the White House concurred.

Companies that agree to the framework will submit their models to U.S. inspectors to be evaluated for safety for 30 days before those companies can receive federal funding—including from the Defense Department, whose 2027 budget request seeks more than $54 billion for AI companies, according to the officials.

A June White House executive order adds that models from participating companies would get extra intellectual property protection from Chinese competitors or others who might seek to steal secrets.

The White House is not commenting on how it will conduct the inspections. One of the officials said the the Office of Science and Technology Policy is still trying to set testing standards and how bodies like the National Institute of Standards and Technology and the Cybersecurity and Infrastructure Security Agency will conduct or design tests.

The murkiness around the policy’s details has angered some top lawmakers. In a Tuesday letter, Senate Democrats ask the White House to “provide an unclassified response, with a classified annex if necessary, clarifying the Administration’s current policy and approach to limiting access to advanced AI models.”

The lawmakers also evinced concern about the models' new abilities to defy easy inspection, scrutiny, and limitation.

“During an internal evaluation [in July] OpenAI models escaped their testing environment and used high-level technical capabilities to compromise a third party’s network without any instructions to take those actions,” the letter reads. “The Federal Government cannot be passive as these capabilities emerge.”

AI’s Jurassic Park moment

Last week, Chinese company Moonshot AI released a new open-weight model, Kimi 3, that performs as well as some top U.S. models and is being offered to consumers around the world at a far lower price.

Lawmakers and others are increasingly worried the United States could fall behind China in a race to develop faster, more efficient, more profitable models and to shape the way global populations use, buy, sell, and even build AI. Open weight models have become a key point of contention.

Anthropic wants more scrutiny of open-weight models and more efforts to curb sales of high-performance chips to China to thwart distillation attacks. But others in the industry are taking a more supportive view of open weights.

A former senior White House official and a former senior defense official with direct knowledge of the discussion said Anthropic pushed for more language in the framework to address open-weight security, but came away disappointed.

Anthropic did not comment for this story.

In recent months, several of the newest AI models have broken out of their virtual testing containers.

Anthropic disclosed the first such incident in April, when an early version of their Mythos model was able to “autonomously write some remarkably sophisticated exploits,” including one that allowed it to escape an isolated testing environment that programmers use to test code for effectiveness and safety before it’s released.

Anthropic pulled the model from general release, but made it available under "Project Glasswing" so the government and a handful of large companies could find and fix vulnerabilities in their software.

The White House responded with an export-control ban on June 12, barring access to the model not just to foreign countries but even foreigners in the United States. That meant that Anthropic’s own researchers, many of whom were born outside of the United States, could not work on the model. The White House reversed the ban on June 30.

In July, tensions around a national AI safety strategy, or lack thereof, grew hotter. Both Anthropic and OpenAI revealed new incidents in which models breached their containment. It came as no surprise to many veteran cybersecurity experts. AI researcher and author Gary Marcus, in 2022, predicted such a possibility on his blog, describing it as an “AI’s Jurassic Park Moment.”

Last week, AWS Chief Security Officer Stephen Schmidt told reporters: “Containers are not security boundaries. I actually have a T-shirt that says that, which I started wearing about three years ago.”

AWS hosts multiple models for users through its Bedrock platform. Schmidt said the AI Mythos era requires a far more vigilant approach to cybersecurity, especially for researchers.

“One of the reasons that we built the virtualization infrastructure for AWS using our own Nitro Hypervisors so many years ago was we realized that containers were not an appropriate security boundary then. The same is true for AI. You cannot use an AI container as a security boundary.”

In March, a group of British researchers calculated the sandbox breakout period for various large language models, which proved very accurate months later.

A follow-on paper published this week by the same group describes how to build better containment environments for the models emerging today.

]]>