Defense One - Science & Tech

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.

]]>

Pentagon counterdrone task force has a new AI-powered marketplace

Lauren C. Williams — Tue, 04 Aug 2026 22:18:13 -0400

The Pentagon’s counterdrone task force awarded Kaizen Laboratories $15 million to build a “two-way” marketplace designed, in part, to make selling tech to foreign militaries easier.

The site was built from a “blank slate” and developed to facilitate the Pentagon’s Foreign Military Sale Fast Lane Initiative, Nikhil Reddy, Kaizen’s co-founder and CEO, told reporters ahead of the announcement.

“We just launched it, and we'll continue to bring on more partner nations, a ton more vendors to the system,” Reddy said.

The AI-backed marketplace has been online for about two weeks, but details of the contract weren’t publicly released until late Monday. The Pentagon awarded Kaizen the contract on May 8 through a rapid acquisition tool called an other transaction agreement.

The new marketplace aims to provide a single point of purchase for the U.S. military and allies and partners, while helping them find the right technology that’s also interoperable. State and local agencies are also being onboarded, Mike Obadal, the Army undersecretary, said at a Center for Strategic and International Studies event Thursday.

“Networked air defense, whether it's counter-UAS or ballistic missile, is critical because everyone has to be able to see the same picture,” which can mean sharing or having “commonality” in information and equipment where possible, Obadal said.

The Defense Department, other federal agencies, and certain “foreign governments are able to go on the counter-UAS marketplace, where we have a number of different industrial options. We've got all kinds of AI workflows in there. You can say: ‘Here's the problem I'm trying to solve, and it'll set up a package for you.’”

The platform is particularly helpful with foreign military sales, for which the portal and its Amazon Web Services backbone allows the Pentagon to “start the paperwork for a foreign military partner that immediately becomes a letter of intent or a letter of request,” Obadal said. “The initial ask from a foreign government is, ‘we would like to investigate these capabilities.’ Traditionally, that goes into our FMS, our foreign military sales process, which takes months at best to get to the other end.”

Australia, Poland, Republic of Korea, Romania, and the United Kingdom are currently approved to use the marketplace, which could help the U.S. align its counterdrone tech with their militaries.

“Allowing them access to the pricing, the availability, the capabilities, all on a digital database … And then the backside of that, generating the FMS process and starting that FMS process is incredibly powerful,” Obadal said. “And so now we can come together with an informed discussion on how do we network counter-UAS between our countries, our deployed forces, our partners…That's what we're trying to promote with the use of modern tools like the marketplace.”

Reddy said he wants to make the marketplace the best “on the internet,” with lots of features, such as integrating test data, that he hopes to add in a matter of weeks.

“There's a ton of agentic tooling that we want to build into the platform, discovery, comparison of products…Testing integration so that you purchase an asset and you can immediately see testing and integration data to ensure that compliance is there,” Reddy said. “We've met with logistics leaders across the military that want repairability information and right-to-repair catalogs built into the marketplace. So all of that is coming.”

]]>

How the UK is helping Ukraine hit Russian targets

Patrick Tucker — Wed, 29 Jul 2026 00:47:09 -0400

A new tech-sharing agreement allowing Ukrainians to make a British-designed radar jammer called Stone Cloak may soon help Ukraine accelerate drone strikes on Russian targets.

Ukraine has been using British-built versions of the digital jammer—about the size of a computer tablet—on long-range strike drones already this year. But the agreement allows the country to mass-produce it themselves so it can be more rapidly scaled across their drone forces.

The British government announced the existence of Stone Cloak for the first time on Saturday as Ukrainian President Volodymyr Zelenskyy met with the new prime minister at a naval base in Portsmouth. The occasion is the annual Sea Breeze joint exercise. Ukrainian, U.S. and British sailors paid “special attention” to countering Russian air defenses during this year’s Sea Breeze according to Ukraine’s national news agency.

“At this point, we don't know the full details of Stone Cloak, but suffice it to say, the system will likely increase mission effectiveness against some of the Russian air defenses,” Jeffrey Fischer, a retired U.S. Air Force officer, told Defense One via email.

Ukraine uses a variety of electronic warfare systems as part of its drone operations. But the ones provided by partners like the U.K. are often backed by larger research budgets. Operators and developers can also test them in a more controlled environment than the Ukrainian battlefield. But they can emulate the unique electronic warfare aspects of such places–so long as Ukrainian operators are allowed to participate, such as the Hedgehog 2025 exercise in Estonia in which the participating Ukrainian team exposed large gaps in NATO drone warfare concepts. That’s why new technology emerging from joint Ukrainian-allied exercises is a big deal. But the actual sharing of intellectual property related to that technology is more rare.

British Prime Minister Andy Burnham on Monday described Stone Cloak as “the best of homegrown British innovation, and proven on the frontline” of Russia’s war on Ukraine. That frontline has moved back toward Moscow, thanks in small part to British-donated Storm Shadow missiles, which Ukraine has been using to hit targets in occupied territory going back to 2023. However, the United Kingdom never transferred the actual design technology to Ukraine, just the missiles themselves.

The new announcement shows how quickly technology sharing between the two countries has evolved in just three years, as well as the speed of Ukrainian weapons development and manufacturing compared to some NATO counterparts.

The United Kingdom spent six years developing and deploying Storm Shadow. Manufacturing relied on specialized European supply chains. French and Italian partners were involved, which made a tech transfer bureaucratically difficult. And the missiles were expensive gifts, averaging $2 million apiece.

These and other factors compelled the U.K. and Ukraine to enter into a joint agreement to develop a new long-range strike drone with streamlined and simplified requirements—a 2024 effort called Breakstop that British officials say will deliver missiles for frontline use later this year. That timeline—less than 24 months from concept to full deployment—is far faster than Storm Shadow and much less expensive. Meanwhile, Ukraine also announced its own long-range strike missile, the Flamingo, or FP-5, which is also capable of striking Russia but costs much less than the Storm Shadow.

Fischer said that the contrast in innovation speeds helps explain how such intellectual-sharing partnerships help the United Kingdom as much as they do Ukraine. “One critical but often overlooked lesson from the Ukraine war is the extremely rapid adaptation cycle Ukraine is able to implement. In some cases, battlefield lessons are implemented in less than 10 days to new systems. In my seven combat deployments, NOTHING has ever come close to that.”

]]>

NIST unveils new AI evaluation platform

Alexandra Kelley — Mon, 27 Jul 2026 16:56:00 -0400

The National Institute of Standards and Technology launched a new program on Monday granting researchers access to an isolated testbed environment to safely evaluate artificial intelligence models against various commands.

The AI Technology Evaluation, or AITE, is a voluntary testing vehicle focused on AI model safety analysis. It provides blind data for models to process when completing tasks to gain objective insights and conduct evaluations of model capabilities. Notably, the evaluation data is not intended to serve as training data for the models.

Initially, AITE will focus on conducting image analysis tasks using large vision language models across three domains: quantum science, genomics and public safety. More tasks will be available in the future.

AITE’s fundamental goal is to offer a universal rubric to effectively evaluate AI models' capabilities and determine the state of the art for model performance.

“The infrastructure provided by NIST will provide common data, metrics and scoring to help developers understand the performance of their models,” the press release said.

Both data providers and model providers working with AITE will need to submit materials related to the testing. Data providers are asked to submit original datasets that are inaccessible publicly, along with a “meaningful” task suited for the data.

Model developers, likewise, will submit their AI models to be tested using the datasets. The first set of evaluations will start in August 2026.

The formation of AITE is the latest step in the Trump administration’s strategy to work with major AI developers in advancing model safety through voluntary model submissions.

The Commerce Department announced a renegotiated deal in May between the agency and three companies—Google Deepmind, Microsoft and xAI—to evaluate their models through the Center for AI Standards and Innovation.

]]>

Lockheed Martin unveils baby Patriot missile to address urgent gap

Patrick Tucker — Mon, 20 Jul 2026 21:29:40 -0400

With demand for Patriot PAC-3 interceptors sharpened by Russian and Iranian ballistic missile attacks, Lockheed Martin unveiled plans to build a smaller and cheaper, if less capable, version of the exquisite interceptor.

The Adapted Capability Effector, or ACE, missile would use the PAC-3's launcher and command and control software, but a less expensive rocket motor, radar, and electronic brain, according to a company announcement and remarks by Lockheed Martin Missiles and Fire Control president Tim Cahil at the Farnborough International Airshow.

The ACE lacks some of the features to counter the most advanced missiles, said Cahil, but is intended to be useful against the type of lower-tech short-range ballistic missiles that Iran and Russia have used in recent destructive attacks.

Said one missile defense analyst, “It is conceivable that if you affect the cost drivers, that you—that's probably how you get down to a lower cost. The cost drivers and the bottlenecks of the current PAC-3 missiles are things like the seeker.”

Tom Karako, the director of the Missile Defense Project at CSIS, said, “The United States and its allies, the entire PAC-3 community needs as many PAC-3s as industry collectively can produce. But we're also going to need alternatives, and not everything needs the most exquisite capability. And so this is why the low-cost interceptor competition is important and urgent.”

Cahill cautioned that the missile is not “designed to go after drones.” It is, however, designed to address a defensive gap that today’s wave of defense tech startups have been unable to fix: a less expensive way to shoot down, say, an Iskander-M traveling at 5,600 miles per hour.

While Ukrainian and U.S. companies have developed counter-drone solutions, short-range missile defense is one of the few areas where Ukraine can’t replace a U.S. capability.

That’s a deadly problem for Kyiv. Last July, the White House claimed that its aid to Ukraine had drawn the Pentagon’s stocks of Patriots. That turned out not to be true at the time, but became so after President Trump launched his war on Iran in February. An By April, CSIS wrote, the United States had used up about half of its supply.

When Ukraine ran out of PAC-3 interceptors this month, Russia responded with increased short-range attacks.

That supply-and-demand gap is prompting more defense companies to try to produce ballistic-missile interceptors. But so far, none has succeeded. The closest European alternative to the Patriot, the French-Italian SAMP/T, is not as effective, Ukrainian officials have said.

Lockheed, meanwhile, is pitching the idea of ACE missiles to at least some of the 18 countries that already have compatible PAC-3 launchers.The company envisions “joint development and production with European partners,” according to their statement.

Cahill said the new interceptor might enter production in 2028—or possibly even sooner, depending on how ready governments are to open their pocketbooks and work together.

]]>

Can new drones, 3D printers defeat distance’s tyranny? RIMPAC aims to find out

Jennifer Hlad — Thu, 16 Jul 2026 16:36:41 -0400

SCHOFIELD BARRACKS, Hawaii—Replacing broken parts and resupplying ships at sea is a challenge anywhere, but in the Pacific theater, that problem is magnified by thousands of miles of sea water. Now, as part of the sprawling Rim of the Pacific, the U.S. military is practicing how it might use tech like self-driving boats and 3D printers that can be thrown out of airplanes to overcome the region’s oft-lamented tyranny of distance.

“We are piloting this program during RIMPAC to experiment with the idea of what theater-wide advanced manufacturing is going to look like for the joint force,” Rear Adm. Michael Mattis told reporters during a media day at the U.S. Pacific Command Joint Advanced Manufacturing Center, calling it the U.S. military’s largest advanced manufacturing demonstration so far.

This year’s RIMPAC features “38 countries, 31 surface vessels, five submarines, over 30,000 personnel, almost 180 aircraft, and over 1,100 personnel that are going to be doing our landings,” Mattis said. “In that process, it’s an enormous experiment for us to figure out how to sustain the joint force.”

Marine Lt. Col. Michael Radigan said the experiment—coordinated by Fleetwerx and the Naval Postgraduate School’s Consortium for Advanced Manufacturing Research and Education—is bringing together a “trifecta of advanced manufacturing, autonomous systems, and artificial intelligence… to deliver true parts that the joint force needs.”

“What are we trying to do? Very simply, this is Uber for manufacturing delivered at the speed of Amazon, for the highest quality of parts, and we have the opportunity to do it in the most contested of environments. And that’s extremely challenging,” Radigan said.

Some of the experiment’s technology was put on display in a sun-scorched field at this Army base near Oahu’s North Shore. San Diego-based Firestorm gave a tour of its expeditionary advanced manufacturing cell, which comprises two shipping containers that can be set up by two people in a few hours. Norway-based Fieldmade displayed what it calls an “additive manufacturing micro factory”—a rugged 3D printer that can be parachuted from a plane and set up in minutes.

Florida-based Snowbird Technologies showed off its own 3D printer, housed in a compact shipping container, that works on land and aboard a ship at sea.

Rhode Island-based Havoc exhibited a 14-foot autonomous boat it printed for the exercise. The solar-rechargeable unmanned surface vessel is built to operate in moderate seas—and to right itself if it capsizes in heavier weather, said Sea Thomas, Indo-Pacific director for the company.

And Los Angeles-based Splash Industries displayed a 10-foot Typhoon, an autonomous drone boat with a Pelican case attached to the top. CEO Ivan Avanesov told reporters that the USV had recently executed the first autonomous resupply of a warship at sea when it drove straight onto the well deck of the USS Essex amphibious assault ship underway during RIMPAC. The Typhoon USV later drove itself through rough waters to an aircraft carrier 100 miles from shore, Avanesov said.

But the gear is just part of the puzzle. “I think we’ve actually got the technology pretty good,” Radigan said. “It’s all the other things that have to happen throughout the theater in order to be able to deliver those [high-quality parts], the logistics, and creating this into the scale that you’re going to see today. So it’s all about operationalizing advanced manufacturing.”

]]>

How a former Marine is rewriting the future of battlefield AI

Patrick Tucker — Tue, 14 Jul 2026 16:11:28 -0400

One stubborn former Marine is changing how the Pentagon moves data to farflung troops—a change intended to enable them to use advanced AI tools, and one that just might reshape how lightweight models are developed.

Some of these new workflows have already been used during the U.S. war on Iran.

“If you look at what's happened with Operation Epic Fury, in particular, we were able to incorporate dozens of new feeds in real time that allow us to not only serve up that data in the right format, right structure, and everything else for those, for those applications to leverage, but also get data at the speed of conflict,” Cameron Stanley, the Pentagon’s Chief Digital and Artificial Intelligence Officer, said at the AWS Summit in Washington, D.C., last week.

Stanley was talking specifically about the War Data Platform, a branch of the Pentagon’s Advana data system being developed to digest thousands of data feeds. The WDP is a key part of defense leaders’ plan to use AI to help commanders make decisions faster.

But another key to the plan is developing ways to get those tools and that data to troops in and around the battlefield, where connectivity is minimal or absent. Nowhere is the challenge more acute than in Indo-Pacom Command, with its vast distances and powerful potential adversary.

That’s the main challenge picked up by Bala Selvam, an entrepreneur and a former Marine, when he stepped into the job of chief technical officer for Special Operations Command Pacific in early 2025.

A race judged in milliseconds

When Selvam took the job, he found that the Defense Department’s data workflows were largely being developed in places with excellent networks and plenty of computing resources. That had produced a culture of shipping huge amounts of data with little concern for latency.

In the continental United States, or even European or Central Command, “It didn't matter because you had all the compute you needed,” Selvam said at the AWS event.

But that wouldn’t work at INDOPACOM.

The distance from the command’s Hawaii headquarters to its second-largest data center is 9,000 miles—too far for efficient data transfer, even with low-earth-orbit satellites, Selvam said.

Analysis and bureaucracy might stretch the data’s journey by yet more thousands of miles before it gets to warfighters at the edge.

“If you're SOCOM, it goes back to Tampa, or if you're any other service, back over here”—that is, Washington, D.C..

All of that was limiting how quickly operators could merge top-secret data and open-source data, analyze it, send it to troops, and share it with partners. The deadline is a matter of milliseconds, with China working 83 percent as quickly. So staying ahead of that number is critical to moving information faster than the adversary. Losing that race just might mean losing a conflict.

The problem only got harder as more data streams emerged. And it also meant that troops in or near conflicts would be unable to use advanced AI tools that require heavy use of enterprise compute resources.

The approach

Rather than try to solve each problem he discovered, Selvam decided to figure out just what everyone actually needed. It’s the kind of analysis that a big U.S. consultancy might require millions of dollars to complete.

Selvam started by going to INDOPACOM senior leadership. “‘Write down everything that you expect everyone to do,’” he told them. "I think a list is going to be in the hundreds. It was actually like 10 things.”

Selvam took this list of “commandments” and worked his way down chains of command, asking everyone to lay out what they were using and how. He found many commanders buying software and tech tools that didn’t match up with what they needed to accomplish.

One category, in particular, stood out.

“They kept saying, ‘the AI.’ I got so mad that I said, ‘If you do it, I might, like, throw a wild dog in your house,” he said. “If you're gonna say AI, tell me what exactly you're doing…If you can’t tell me that, you're not using AI.”

To force commanders to define their tasks, Selvam began to physically deprive them of their technology.

“I actually almost got, like, fired. I took everyone's computer away, and then I hid it. Then I put whiteboards in everyone's rooms with their actual task. I forced everyone to walk into the [Joint Operations Center] to put in their inputs. When they said the words, ‘I wish I had this technology; I did this virtually,’ I gave them back the computer with only that thing. It was like an Excel sheet and a PowerPoint slide. And then when they said, ‘Man, I wish I could do this faster,’ then I gave them back Palantir and Databricks [a cloud-based data analysis and AI platform] and everything else.”

For Selvam, this wasn’t about instilling tech discipline; it was implementing business best practices. The process yielded a record of what everyone actually needed to do their jobs.

“If you want to know what a good tech company looks like, look at their documentation. Same concept here,” he said. “So, for every bit of our work, we will write processes, procedure business rules, and we will build based on those process procedures or business rules.”

A key meeting

Armed with a much better understanding of what people actually needed to get done, Selvam turned his attention to the problem of getting data where it needed to go. For that, he needed help.

Around January, he traveled to the Crystal City, Virginia, offices of AWS, which makes billions of dollars selling the Pentagon the sort of enterprise cloud capability that won’t be available at the edge of combat. There he met Joshua Hobgood, who leads AWS’ efforts to sell AI-related services to the world’s militaries.

“I want to solve for AI at the edge,” Selvam told Hobgood, who described the meeting to the AWS Summit audience. Hobgood answered that it was a common problem, to which Selvam replied, “Yeah. But we're going to do it right.”

Over the next two hours, they hammered out a new plan. First, they would build smaller cloud compute clusters, or nodes, that would require less power and infrastructure, and, hopefully, be harder to target in a conflict. These would be designed to support battalions—a few thousand troops—and would replace larger nodes meant to cover entire divisions of 20,000 or more, Hobgood said.

He said that making this work will require commanders to follow Selvam’s procedures governing which tools people use and for what purpose. When SOF commanders “give out tasks, they will specifically say, ‘Humans, you will do these [six] out of these 10 tasks…These other four, here's how you're gonna let the [AI] agents take over that.’”

The next step was to make sure those agents, or whatever other AI tools were running at the edge, were feeding data back to the cloud.

“We based the solution on the idea that…up to the point of a conflict scenario, there will be connectivity. So we wanted to create a system that could continuously learn,” Hobgood said. “The core assumption to that was that the majority of your AI inference, your data analysis, would happen in the cloud, in the AWS cloud.”

But as they worked to flesh out the ideas over the next few months, Selvam and Hobgood realized that AI tools could be built to sip power and compute—and still deliver results to troops in the field.

“We're at a point where we can use smaller and smaller models…to get, actually get, the best results,” he said.

That’s going to be vital as the Pentagon works to integrate more and more information into the War Data Platform. In May, the Chief Digital and Artificial Intelligence Office said the WDP was drawing upon some 4,000 data sources from more than 55 organizations.

“What we're trying to do is really create that single pane of our single repository of truth for a lot of data that was disaggregated and come together into a single platform that allows for very quick connection of data sources to different types of applications,” said Stanley, who leads the CDAO.

It’s unclear how Selvam’s proposed brigade-sized compute clusters will relate to the platform.

But, said Stanley, “That's going to become a program of record here pretty soon.”

Selvam said the effort was supremely worth it.

“I'm not gonna lie to you; my entire 18 months has been like, ‘If I don't solve the [cross-domain solution] problem…my life would just be meaningless.’ But if I did solve it, I'm going to put it on my tombstone…It's going to say, ‘I solved the [Department of War] CDS problem.’ That's what it's going to say, I swear to God.”

Nextgov’s Alexandra Kelly contributed to this story.

]]>

AI can now power every stage of a cyberattack

David DiMolfetta — Mon, 13 Jul 2026 21:00:00 -0400

Just two years ago, hackers were tapping into generative artificial intelligence to probe targets, translate technical material and troubleshoot malicious code. The technology sped up some key parts of a cyber operation, but other stages remained solely in human hands.

That line is now beginning to blur. In a range of cyberattacks observed over the past year, AI systems generated commands, tested vulnerabilities, and helped hackers move through victim networks, sometimes carrying out thousands of commands with less human direction than researchers had previously seen, according to research released Monday night by cybersecurity firm Check Point.

It means AI has now been used in some form at every stage of a cyberattack, from identifying targets to exploiting vulnerabilities to stealing data, to help hackers achieve their goals.

The shift does not yet amount to fully autonomous hacking, a top company executive told Nextgov/FCW, but it shows that AI is moving from an occasional aid to an extra set of hands throughout the entirety of an intrusion process. The findings also highlight how rapidly the global cyber ecosystem has adopted AI, with the technology now embedded across every part of an offensive operation rather than confined to isolated tasks.

“We watched criminal groups breach government agencies at scale, using AI as the primary operator rather than a background assistant,” the Check Point report says. In most cases, the AI model’s role was revealed by the attacker’s own mistakes or monitoring by the AI provider, rather than tools or safeguards deployed by the victim organization.

For these AI-enabled pursuits, hackers have used open-source models and purpose-built malicious AI tools sold on the dark web, but major commercial providers remain their primary choice, company threat intelligence lead Sergey Shykevich said in an interview.

The research points to the Gentlemen ransomware group as one instance of how AI is being folded into routine criminal operations. Members compared mainstream commercial models based largely on which imposed the fewest restrictions and used AI to help build internal tools, including a management platform developed in three days.

The findings also highlight VoidLink—a sophisticated toolkit for remotely controlling infected computers — that researchers initially believed had taken a team several months to build. Check Point later found that a single developer produced roughly 88,000 lines of working code in under a week using a commercial AI coding tool.

Hackers tend to first choose U.S. AI models like ChatGPT or Claude because their responses are generally considered higher-quality, but they have a tougher time exploiting those families of models because of stronger guardrails designed to prevent malicious use of the tools, Shykevich said. When those attempts fail, they then pivot to Chinese-made AI platforms that have lower guardrails.

“They are trying to jailbreak those [Western] models, but when they are not successful, they just go to DeepSeek, Qwen and Trae,” he said, referring to a trio of AI models and platforms originating in China. Qwen is a suite of AI and large language models developed by Alibaba, while Trae is an AI-backed code editor and programming platform built by ByteDance. Ransomware gangs have been heavily leaning on those Chinese models to help generate code for their exploits, he noted.

Chinese AI models have recently grown more capable and widely used for coding tasks. Beijing has discussed restricting overseas access to some advanced models, underscoring their growing national security value and their appeal to cybercriminals seeking tools to help with their exploits.

When asked how these new dynamics involve targeting U.S. government and critical infrastructure organizations, Shykevich said there are certainly higher and faster levels of exploitation attempts, though he assessed that’s due to a combination of AI-enabled speed and geopolitical tensions.

He said that, in an ideal world, security flaws should now be patched within several hours of discovery, though he immediately acknowledged that would be next to impossible. The Cybersecurity and Infrastructure Security Agency recently revamped its remediation timeline guidance, ranging from three days for the highest-risk flaws to 60 days for lower-priority issues.

The move is part of CISA’s response “to the current threat landscape where AI software services can assist threat actors to find and exploit vulnerabilities,” the agency said last month.

The findings come as the latest tranche of frontier AI models is showing sharp gains in their ability to perform cybersecurity tasks. OpenAI last week released GPT-5.6, which it called its strongest cybersecurity model yet, reporting substantial gains over its predecessor on benchmarks testing exploit development and proof-of-concept generation.

The Trump administration, meanwhile, has given the National Security Agency, CISA and other federal officials until Aug. 1 to develop a classified process for benchmarking the advanced cyber capabilities of frontier AI models and determining which systems need additional government scrutiny.

“The most significant shift this report documents is not a new technique. It is pace,” the Check Point report says. “A vulnerability now becomes a working exploit within hours of disclosure. Phishing campaigns run at a quality and volume no human team could match. Intrusions span dozens of targets simultaneously, with AI handling the operational work between check-ins. Security teams working at human speed cannot match that cadence.”

]]>

How Ukraine won the first great robot war

Patrick Tucker — Wed, 08 Jul 2026 17:02:31 -0400

In the spring and summer of 2026, the story about Russia's war in Ukraine changed. The defenders are combining air and ground drones, autonomy, and swarming to re-capture territory with robots while risking fewer humans—increasing casualties for Russia while decreasing them for their own forces. Some four and a half years after confidently launching his invasion, Russian President Vladimir Putin has dwindling military options.

In this short documentary, analysts, entrepreneurs, European experts, and NATO military officials describe how Ukraine defied expectations and changed the future direction of military technology—and of war itself.

]]>

Lightweight cells powered by nuclear waste could drive tomorrow's drones

Lauren C. Williams — Fri, 03 Jul 2026 08:00:00 -0400

How many missions could a drone or satellite fly with a battery pack that can last decades? And what if that battery could be fueled by nuclear waste?

That’s the future scientists are working toward in DARPA’s “Rads to Watts” program, which aims to create lightweight batteries with a high energy density. And a recent $3.37 million contract award aims to fund a viable proof-of-concept device that can produce more than 10 watts per kilogram with a yearslong shelf life.

“Solar cells directly convert sunlight into electricity…Ours directly convert radiation into electricity,” said Stafford Sheehan, CEO and founder of Project Omega, which describes their radioisotope power sources as mini-generators that replace traditional batteries.

“We already have some of these small devices running; the ones that are specifically designed to meet the DARPA figure of merit are going to come out early next year.”

Several organizations are participating in the program, with Morgan State University serving as the prime contractor and handling basic research and the Pacific Northwest National Laboratory handling nuclear materials and testing. Northrop Grumman and ARA will bring computational modeling to make sure the prototype meets performance standards.

Project Omega will build the nuclear power generator based on a radioisotope found in nuclear waste, and Widetronix is designing the semiconductor power converter. The goal is to produce a working prototype by early 2027 at the Pacific Northwest National Laboratory.

The power cells could be used in “any application where a battery dying is a pain point,” Sheehan said. “One example is on satellites: if you lose power on a satellite, you lose the satellite, it's gone…if your batteries die and you don't have any sort of backup power.”

These power sources use isotopes separated from nuclear waste and convert radiation directly into electricity.

“At a high level, we take nuclear waste, we recycle it into two products: one is fuel for reactors…the other are power isotopes, so isotopes you can use to power things,” Sheehan said.

Radioisotope power sources have been used in everything from smoke detectors to space systems. But Project Omega hopes to do it on a larger scale.

“There are over 100,000 metric tons of nuclear waste sitting in the 52 reactor sites around the country; so there's plenty of nuclear waste currently. The federal government gets sued for billions of dollars every year just because they haven't dealt with the nuclear waste,” Sheehan said. “It's very valuable to have a battery that lasts.”

Omega’s power cells consist of a solid state, or “chunk,” of isotope that will be layered with the semiconductor to generate power. They also work in extreme temperatures—something that would benefit military operations using unmanned systems in harsh environments.

“We have been using these radioisotope power systems in space for decades,” Sheehan said. “We're just taking the systems that we use for space and we're using a different isotope,” Strontinum-90, which is less hazardous than the Plutonium-238 isotopes used in similar systems.

The award comes as the Pentagon grapples with increased demand and use of drone systems

that have to be charged and the persistent need for more power generation on the battlefield.

“Over the next 18 months, the program will focus on reducing technical risk, testing system performance under realistic conditions, and generating the data needed to inform future development and transition pathways,” a PNNL official wrote in a statement to Defense One. “Key challenges include improving energy conversion efficiency, validating long-term reliability, managing radiation effects, and ensuring safe, secure handling and deployment.”

]]>

GenAI.mil records almost 1.7M users, plans new model additions

Alexandra Kelley — Thu, 02 Jul 2026 09:00:00 -0400

The Department of Defense plans to bring new models onto its internal, department-wide artificial intelligence marketplace and deploy them at higher classification levels, part of its updated procurement policy that aims for “commercial-first” in its deliverables.

Cameron Stanley, the chief digital and artificial intelligence officer at DOD told attendees at the AWS Summit in Washington, D.C. on Tuesday that as GenAI.mil reached a record 1.7 million users—along with the creation of over 100,000 custom agents—even more models will soon be made available on the platform.

“We're looking forward to advancing, getting new models on to GenAI.mil, we're looking at GenAI.mil going to higher classification levels,” Stanley said. “It's just a really exciting time for generative AI in the department.”

GenAI.mil already hosts capabilities from SpaceX, OpenAI, Google, NVIDIA, Reflection, Microsoft, Oracle, and Amazon Web Services that are available at Impact Level 6 and 7, as the Pentagon announced in May.

OpenAI confirmed in mid-June that its flagship chatbot, ChatGPT, will be eligible for controlled, unclassified information through GenAI.mil in July.

One major use of AI for the DOD has been to aggregate data for warfighters, resulting in faster decision-making. While Stanley said that “well-trained” soldiers, sailors, airmen, guardians and marines traditionally make these critical military decisions, AI is helpful in parsing through large volumes of data quickly

“Human cognition is just not going to be able to keep up in a lot of battlefields,” he said. “So what we've done—very successfully—is identify ways where we can accelerate certain identification of the right pieces of data or information in order to make that better decision.”

Stanley clarified that the addition of agentic tools to support analytics is deliberate, and features “very tight guardrails” to accelerate the analyses that he estimates would take two to three human analysts operating in disparate systems to identify.

“So we go from—instead of having six or seven systems we have to go across in order to make that decision—we're now doing it instantaneously, or nearly instantaneously, with humans appropriately managing the entire workload process and actioning it from the same system that we identified the decision from,” he said.

The goal of Stanley’s office, he said, is to be a “commercial-first organization.”

“We're trying to put the vendor next to the warfighter and have the vendor have one goal, one job, that's it, and that is to deliver exactly what the warfighter’s needs are,” Stanley said. “What we do is we create the environment with the right tools and the right environments with the right security in place with the right contracts in place.”

]]>

Pentagon recruiting new tech talent for AI implementation

Patrick Tucker — Tue, 30 Jun 2026 10:04:38 -0400

The Pentagon on Tuesday announced a new campaign to recruit hundreds of young programmers for AI and engineering challenges, according to an advance notice seen by Defense One.

“War Force is a call to action for patriotic forward-deployed engineers who want to serve their country and the warfighter,” Defense Department CTO Emil Michael said in a statement viewed by Defense One.

Successful applicants, referred to in this OPM job listing for a “forward deployed engineer(s),” will work for two-year stints to implement the department’s AI Acceleration Strategy as well as other critical IT needs. Most of the positions will be based in Washington, D.C., and candidates should be able to hold a top-secret clearance.

The “War Force” is a part of a broader push between the White House and participating tech companies to bring early-stage software talent into the federal government, an effort dubbed “Tech Force.”

Those selected to participate will work with a variety of teams on implementing frontier AI, machine learning, automation, scaling new software tools and solving other critical IT needs across the military and the Department.

War Force recruits will “join a cohort of individuals that receive leadership training, get exposure to America's top technology CEOs and gain unparalleled exposure to policymaking and national-scale impact,” according to the listing. “This may include CEO fireside chats, networking events, coding training/certificates, and other opportunities.”

]]>

Pentagon launches ‘War Force’ initiative to onboard tech talent

Edward Graham — Tue, 30 Jun 2026 09:45:00 -0400

The Office of Personnel Management and the Pentagon jointly announced the launch of a new initiative on Tuesday to bring top software engineering talent into the Department of Defense, part of a broader governmentwide push to hire more skilled technologists.

The recruitment effort, called War Force, operates under OPM’s larger Tech Force program. That initiative launched in December 2025 to onboard tech and cybersecurity professionals across federal agencies, although it notably followed Trump administration moves to let go of thousands of tech-focused workers and shutter several innovation-focused units.

OPM Director Scott Kupor said in a statement that War Force “builds on the momentum of Tech Force by connecting outstanding engineers with opportunities to solve complex challenges alongside the War Department.”

President Donald Trump signed an executive order last September that authorized DOD to use the “secondary title” of War Department — framing that is also reflected in the initiative's name.

OPM said the War Force recruitment effort “will launch with a targeted hiring campaign,” with a focus on bringing in applicants who have experience “deploying and integrating advanced technologies — including frontier AI, machine learning, automation, and data systems — and designing, building, and maintaining reliable software solutions that directly support operational needs on behalf of the American warfighter.” Job applications will be accepted through July 10.

Pentagon Chief Technology Officer Emil Michael said in a statement that the War Force will help with “executing the key tenets of the War Department’s AI Acceleration Strategy.” That guidance was released in January and outlines the department’s plans for rapidly integrating AI capabilities into its operations.

The War Force recruitment effort also comes after DOD announced in April that it was launching a cyber apprenticeship program this summer to bring more skilled personnel into the agency.

During public remarks at the SAP NOW summit in Washington, D.C., last week, Pentagon Chief Information Officer Kirsten Davies said the apprenticeship program “has already generated more than 70,000 inquiries,” despite the effort not officially launching until July.

]]>

Army using AI, robot boats for Pacific logistics

Jennifer Hlad — Tue, 30 Jun 2026 07:22:25 -0400

The Army’s 8th Theater Sustainment Command is using artificial intelligence “to help us make better-informed decisions” for supply chain management in the Pacific, the unit’s commander said Friday.

“For logistics, a lot of what we do is very similar to what the commercial world does, and so I have leveraged, and we are leveraging commercial partners with, you know, how do they do warehouse management regionally, and then how do they look at, how do you time delivery of supplies to the location it’s needed, and kind of, what are those time-distance factors,” Maj. Gen. Gavin Gardner told reporters. “ I’m looking at partners, and I’m talking to partners that do that on a global scale, because the distances between the continental United States to the forward positions that we train or live at, like the Republic of Korea or Japan, we’re constantly looking for smarter ways to do that.”

“When resources are not unlimited, how do you best look at demand analysis over time and space, and stock forward the right things versus stocking everything? Because we just can’t afford to do that,” he said. “And so we’re using AI right now to help us see that.”

The command is also already using “very capable” autonomous watercraft in the Pacific and is working with industry to develop and test larger and faster vessels, Gardner said.

“We have industry partners that are creating vessels right now in the water that are over 100 feet long that would move anywhere between four and eight 20-foot equivalent units… think the containers you see moving up and down most of your ports today,” he said. “That capability is already in play, and it’s going through experimentation now, and the 8th TSC, along with the rest of the United States Army, is looking to partner, test, and innovate those autonomous watercraft first out in our region, because if you can work in the Pacific, you can work anywhere in the world.”

But the rules will need to evolve to keep up. Current U.S. maritime laws require a minimum crew size for vessels, limiting autonomous operations to testing or pilot programs for commercial use.

“We need to get the laws of the sea that our Coast Guard currently use to mandate watercraft operations that currently have to be manned when it enters into a port, that we’re comfortable having unmanned systems enter into ports, so that we can rapidly receive and download autonomous watercraft like we do manned watercraft today,” Gardner added.

Gardner said he’d also eventually like to see “anywhere between 30 and 100” medium-sized autonomous vessels in the U.S. Indo-Pacific theater, “berthed up everywhere from Korea to Japan, the Philippines, Australia, Singapore, Thailand,” to help meet the “constant demand for Army watercraft time now to deliver equipment and supplies in the theater.”

The unit has one Maneuver Support Vessel (Light), a manned landing craft that will eventually replace the Vietnam-era Landing Craft Mechanized-8, and has been testing it for about nine months to get data to shape the future design of the vessel. The vessel is nearly four times faster than the older watercraft, and much smaller—but also “much more capable” of operating in shallower water, Gardner said.

“So I’m actually going to different sets of beaches, and so you can put up to two HIMARS onto an MSV-Light, approach a beach very rapidly, bring it into position, download it, and then those platforms can get off and do their fire missions,” he said. “We’ve rehearsed this quite a bit with both the 25th Infantry Division and the 3rd Multi-Domain Task Force,” and have also put multiple Marine NMESIS anti-ship systems onto the vessel for “rapid insertions.”

]]>

Agentic-AI tools aim to give US commanders new target options ‘within seconds’

Patrick Tucker — Sun, 28 Jun 2026 08:00:00 -0400

A new agentic-AI tool set will continuously scan intelligence feeds and operational networks to provide U.S. military commanders with targeting options “within seconds,” the Pentagon announced Thursday.

Dubbed Agent Network, the new tools will employ “agents”—artificial-intelligence entities that perform tasks on behalf of a user, such as running a scheduled search or executing an email campaign—to “continuously scan defense intelligence and operational systems, translating findings into clearly presented options,” said a press release, which added: “Agent Network does not autonomously select or strike targets; it ensures commanders remain in charge of every decision.”

The network is one of seven “pace-setting” projects originally unveiled in January along with a new Pentagon AI strategy. Key contractors in the Agent Network effort include Lumbra and Palantir, which already handles much targeting analysis through its Maven Smart Systems contract.

But expectations for what agents can currently do may be running ahead of reality. “Tasks that AI agents are instructed to perform can clearly have computational complexity beyond” what current large language model architectures can handle, Vishal Sikka—a former CEO of SAP—wrote last July.

Citing the seminal Time-Hierarchy Theorem, Sikka noted that transformer models approach difficult tasks and simple ones using the same mechanical formula. These models can only perform so many operations per “token,” which is the way large language models understand word concepts. Even dealing with seemingly simple concepts can require a large number of tokens. Because of this limitation, there is no way to get a transformer-based model to not hallucinate when the task that you give it is more complex than it has tokens to bring to that task.

“Despite their obvious power and applicability in various domains, extreme care must be used before applying LLMs to problems or use cases that require accuracy, or solving problems of non-trivial complexity,” Sikka concluded.

But Illia Pashkov, founder of SINT Labs and editor of The Agent Times, cautioned against underestimating agents’ potential.

“Agentic AI quietly stopped being a demo this year,” Pashkov said. “It's drafting code, clearing support queues, grinding through back-office work in finance and healthcare, and now it's reading intelligence. The speed is not hype. I've watched these systems compress weeks of analyst work into an afternoon.”

But their capabilities also bring risks—more than people accustomed to working with common AI chatbots might realize. Private-sector companies that have rushed to put AI agents to work are already seeing problems, Pashkov said, pointing to a company whose agent wiped a live production database. Unless carefully implemented, agents can’t tell when they’re going wrong.

“The danger was never a dumb agent; it's a confident one running without a leash, a logbook, or a human who owns the call,” he said.

A host of Defense Department offices and teams are beginning to deploy agent systems, said one DOD intelligence security official who is not directly affiliated with the Agent Network program. The official described an atmosphere of enthusiasm.

“There are so many opportunities to leverage the DOD Enterprise capabilities and allow people to build their own agents,” they said.

But the official allowed that keeping track of how every agent is performing is a major challenge. Governing all of them will be nearly impossible.

]]>

Could the Army’s light squad vehicle power battlefield drones?

Lauren C. Williams — Fri, 26 Jun 2026 16:59:14 -0400

The Army’s proliferating drone use is exacerbating the modern problem of keeping everything charged. One combat team is testing ways to use its light Infantry Squad Vehicles as mobile charging stations.

“As we field technology, power generation becomes increasingly problematic. The [Integrated Tactical Network], all the soldier-borne equipment, all require batteries that have to be recharged. [Enhanced Night Vision Goggles] require batteries, particularly to use thermals,” said Col. Ryan Bell, commander of the 3rd Mobile Brigade Combat Team in the 101st Airborne Division, of lessons from an April training rotation at the Joint Readiness Training Center at Fort Polk, Louisiana.

Finding ways to boost battery power on the battlefield—and learning how long new tech lasts in different conditions—is a persistent problem for the military.

“As we add drones…the [Short Range Reconnaissance], the [Medium-Range Reconnaissance], they all have to be charged,” Bell told reporters Thursday.

So the unit equipped its ISVs with inverters, similar to the kind of thing any car owner might plug into the outlet formerly known as a cigarette lighter to get standard 120V power.

“Every squad functionally had their own generator without having to tow a trailer, or a generator, behind it,” Bell said.

But more power is needed, he said, like hybrid generators that could deliver up to 10 kilowatts, depending on a unit’s needs.

Bell suggested sizing the generators so they can be mounted on the five-seater ISV-Utility variant the Army plans to field. They could also be towed, but that’s not a great solution for rough terrain.

The Senate’s version of the 2027 defense policy bill would greenlight multiyear procurement of the ISVs if adopted in the final version.

“As future ISVs are fielded, we need to ensure they are utilized as power generation platforms, particularly for our squads. Our use of the vehicle as a sustainment platform by adding commercial off-the-shelf racks on the back was very helpful,” Bell said.

“We are also excited to get the ISV-U, which will give us more utility lift capacity. Having a single, common platform across the formation will simplify our logistics chain. The utility variant will be extremely helpful for mission command and power generation for our command posts.”

Bell also said troops also need batteries that carry more charge and last longer, which would “reduce the number of batteries you have to charge and have to carry, because solder load is a real thing.”

]]>

RIMPAC kicks off in Hawaii with a focus on experimentation

Jennifer Hlad — Thu, 25 Jun 2026 04:07:20 -0400

JOINT BASE PEARL HARBOR-HICKAM, Hawaii—The 30th iteration of RIMPAC will feature 30 to 35 experiments that involve unmanned systems, the Pacific Fleet’s second-in-command said Wednesday.

The experiments are “a major part” of this year’s edition of the biennial Exercise Rim of the Pacific, Vice Adm. Jeffrey Jablon told reporters at a press conference. He declined to provide specifics.

Along with the drones, RIMPAC also includes 30 countries, 31 surface ships, five submarines, and nearly 200 aircraft, Jablon said. While the theme is “partners: integrated and prepared,” he said his top priorities for each of the 30,000 participants are “safety, environmental stewardship, and professionalism.”

Though deterring China while getting ready to defend against a potential attack is a major focus for U.S. Indo-Pacific Command, Jablon said RIMPAC “is not about any one particular country or a deterrent for any one particular country.”

And the war with Iran “had no impact” on RIMPAC, Jablon said. “The United States is contributing the same number of forces that we normally contribute.”

Jablon is serving as the commander of the combined task force for RIMPAC. Underscoring the partnership aspect of the exercise, the deputy commander is a Chilean navy officer, the vice commander is a Japan Maritime Self-Defense Force officer, the maritime component commander is a Korean navy officer, and the air component commander is a Royal Canadian Navy officer.

The exercise will end with the sinking of two decommissioned U.S. Navy ships.

]]>

Executive orders seek to hasten quantum computing—and guard against its use

Alexandra Kelley — Mon, 22 Jun 2026 18:06:00 -0400

Two two executive orders on Monday aim to hasten the arrival of useful quantum computers while protecting U.S. systems from them.

“Innovation and security have to be balanced,” National Cyber Director Sean Cairncross said during the White House signing ceremony on Monday.

As expected, the orders address different aspects of the burgeoning quantum information sciences and technology landscape.

One, “Ushering In The Next Frontier Of Quantum Innovation,” launches “a national effort” to create a quantum computer that can perform basic operations and improve quantum sensors. It directs the creation of a "Quantum Computer for Application Development and Discovery Science Effort" at in a Department of Energy facility, while other provisions support quantum computing supply chains, foster workforce development, and explore private sector and international partnerships. The order also expands the Quantum Information Science and Technology Counterintelligence Protection Team to study threats to domestic quantum-computing efforts.

The other order, Securing the Nation Against Advanced Cryptographic Attacks,” pushes the government to move to cryptographic standards that can withstand quantum computers. It puts several agencies in charge of the effort, including the Office of Management and Budget, the Department of Commerce, the Department of Homeland Security, the Cybersecurity Infrastructure Security Agency, and the National Security Agency. The National Institute of Standards and Technology has helmed the effort to identify and test new encryption algorithms.

The order sets a deadline of 2030 to update key elements of critical infrastructure, and of 2031 for "high-impact environments."

Garfield Jones, the executive vice president of Strategy and Research at QuSecure, said the executive order on post-quantum cryptography is an "unambiguous signal" of the need.

"The 2030 deadline for key establishment is a tangible compliance deadline, and the gap between where most organizations are today and where they need to be is significant,” Jones said in a statement. “Agencies and contractors that haven't started a cryptographic inventory are already behind. The organizations that move now will have options. The ones that wait will find themselves managing a crisis.”

IBM CEO Arvind Krishna, who was present at the signing ceremony, said his company “applauds” the Trump administration for both orders.

“Sound policy, sustained investment and public-private partnership are vital to sustaining U.S. quantum leadership and technological resilience,” Krishna said in a statement. “Today's Executive Orders bring that same spirit of policy and investment working in lockstep to the national stage.”

Energy Secretary Chris Wright said quantum computing will join artificial intelligence and advanced semiconductors as the three-part foundation of future computing technology.

“This is tricky. We're not there yet. We're close, but with this executive order and this coordinated effort, we will have scientifically relevant — meaning error-corrected — quantum computing during this administration. The impacts of it will be tremendous,” Wright said during the ceremony.

]]>

CISA now has full Mythos Preview access, people familiar say

David DiMolfetta — Wed, 17 Jun 2026 17:21:00 -0400

The Cybersecurity and Infrastructure Security Agency now has full access to Anthropic’s flagship Mythos Preview model, according to a U.S. official and a second person familiar with the matter.

The cyberdefense agency received access around a week ago, the official said. Both sources spoke on the condition of anonymity to discuss internal deliberations.

The White House Office of the National Cyber Director has not yet set clear parameters for how the agency should use the model, the official added.

The lack of parameters echoes earlier Nextgov/FCW reporting showing federal tech leaders have privately complained that ONCD has not adequately briefed them on implementing or using the model for vulnerability scanning.

CISA did not respond to a request for comment.

Over the last few months, Anthropic surgically rolled out Mythos Preview to select organizations and recently expanded this effort — dubbed Project Glasswing — to partners in industry and other nations. The model has been distributed through a non-public process on grounds that, in the wrong hands, it can significantly boost adversaries’ hacking capabilities.

CISA was not included in an initial Mythos rollout, Axios reported in April. Last week, Nextgov/FCW reported that agency access to the model was imminent.

Mythos Preview is different from Anthropic’s similar-sounding Mythos 5 successor model, which the U.S. effectively banned over the weekend via an export control mechanism alongside the AI company’s Fable 5 model. The move has caused uproar across the cyber and AI community.

Both Mythos 5 and Mythos Preview have only been made available to vetted providers via Project Glasswing.

The Trump administration’s approach to AI has shifted in recent months as officials confront an emerging class of models that can rapidly identify vulnerabilities across computer networks, becoming a major driver of discussions over how AI systems could reshape the future of cybersecurity.

Models like Mythos can help federal agencies identify vulnerabilities faster by analyzing large amounts of software and system data, then surfacing weaknesses and possible attack paths for human defenders to review. Conversely, cyber operators in the intelligence community and Defense Department can also use such models to accelerate their offensive hacking operations.

]]>

AI is taking some parts of background checks from 'months to hours,' clearance agency says

David DiMolfetta — Tue, 16 Jun 2026 13:56:00 -0400

The nation’s largest counterintelligence unit aims to use artificial intelligence tools to speed security clearance reviews for people and companies seeking to do sensitive work on behalf of the government.

The Defense Counterintelligence and Security Agency can use AI to reduce parts of the vetting process from “months to hours,” said Mark Nehmer, an agency analytics and innovation chief who spoke Tuesday on a panel at the Defense One Tech Summit in Virginia.

DCSA is the Defense Department’s main agency for conducting background investigations and vetting personnel for access to classified information, and serves as a key determinant for whether companies are eligible to work with military and intelligence agencies.

A recent congressionally-approved acquisition overhaul, which encourages defense officials to prioritize goods and services from the commercial market, means that the counterintelligence agency will have to process some 43,000 clearance requests per year, he estimated.

“We’re trying to use AI exquisitely, use AI to make these little tiny decisions, and then bring that up to a human, so they can actually have a package of evidence to say, ‘I asked, and this is exactly the conclusion I will come to as a senior analyst that has to make those decisions day-in and day-out,’” Nehmer said.

He did not specify what AI systems would be used for the efforts.

The remarks highlight how the government is applying AI to a key national security function that determines who has access to clearances, and they add another case to a long list of examples showing how the federal enterprise is using AI to speed operations.

DCSA has led the government’s background check process since 2019, when the Office of Personnel Management handed off its National Background Investigations Bureau to the Pentagon.

DCSA’s use of AI builds on a years-long effort to automate and overhaul the federal background-check system. The agency has enrolled millions of clearance holders in continuous vetting under an initiative known as Trusted Workforce 2.0, though the broader modernization effort has faced repeated delays, cost overruns and congressional scrutiny.

Over the weekend, the U.S. invoked an export-control mechanism to essentially ban two major Anthropic frontier models, escalating debates over how Washington could exert itself over AI usage in the government. The decision has been widely criticized.

GovExec Editor-in-Chief Frank Konkel contributed to this report.

]]>

Shaped charges from coffee grounds? Pentagon science chief describes future of war

Bradley Peniston — Tue, 16 Jun 2026 11:21:19 -0400

When the Pentagon’s science and technology chief looks at Ukraine, he sees a war fought with weapons invented, produced, and fielded since the conflict erupted.

“The fact that you can bring relevant capability to the fight, as the Ukrainians and allies have done in the conflict with Russia, that essentially didn't exist at the beginning of the fight,” Joseph Jewell, assistant defense secretary for science and technology, said Tuesday at the Defense One Tech Summit in Arlington, Virginia. “That's the new thing here.”

It’s a thing the United States must learn to do, Jewell said.

Ukraine’s homegrown drone industry “to a large extent, sprung up almost overnight because of urgency. I think with our industrial resources, we certainly could do things at that scale and even in a more sophisticated way. And we need to do it,” he said.

Jewell noted that Ukraine has taken the Russian Navy out of the fight without much of a navy of its own.

“The way they were able to do that, well, there are several things. First of all, their weapon systems were small, relatively undetectable. Second of all, they had a lot of them,” he said.

There is still a need for expensive, highly capable weapons, Jewell said, “But the exquisite effect may be helped along by leveraging a hundred or a thousand drones controlled by AI. And I think that's what we're starting to see modern warfare evolve into. Now, of course, the model is a lot of people in Ukraine who are actually manually controlling these first-person drones. I think the natural evolution of that is AI-controlled or AI-enabled.”

Patent holiday

One way the Pentagon is trying to speed up innovation is by making it easier for defense companies to use government-held technological patents.

The Defense Department holds tens of thousands of patents, but only takes in about $20 million a year from them. In January, Jewell’s boss, Defense Undersecretary for Research and Engineering Emil Michael, announced a “patent holiday” under which private companies can license some of those patents—about 500, Jewell said—free of charge.

He said the first no-fee patent license was granted last month. As of mid-June, 14 patents have been “signed out” for commercial use, one has been licensed for a fee by a company that wanted exclusivity, 36 more are pending, and 145 more applications have come in, he said.

Biotech

Jewell lauded the promise of biotech combined with AI. He cited a new bioengineered thermal coating that may help drones obscure their heat signatures, developed through BioMADE, a DOD-sponsored Manufacturing Innovation Institute.

He also described an experiment in which Marines in the Pacific used 3D printers and other tools to field-produce shaped charges with local materials: plastic water bottles, crushed volcanic rock, coconut husks, and coffee grounds.

“They all detonated, actually; the volcanic rocks were most effective,” he said. “The thing that's amazing to me is this was 3D-printed. You effectively have 99% reduced the time to point-of-use, because you could make it in the field from materials that are endemic in the Indo-Pacific.”

What’s more, the Marines’ shaped charge “had 25% better focusing characteristics than conventionally manufactured high explosives,” he said. “So we envision a future where you have a containerized production facility for potentially the ingredients for that, potentially including the 3D printer to pump out the shaped charges. And then you can drop, say, a CONEX box in the field where you need, so it can produce biodiesel, it can produce jet fuel.”

]]>

Desert e-bike race ‘the perfect’ place to test military-vehicle AI

Patrick Tucker — Tue, 02 Jun 2026 01:29:59 -0400

Pilot Racing, a team in the upcoming Baja 1000 dirt-bike race will bring a special advantage: AI that prescribes when a rider should pit, long before the need becomes obvious.

The thousand-mile trek through California and Mexico is “the perfect” environment to test GDIT’s logistics-and-maintenance AI before it heads off to rough and disconnected battlefields, a company representative said.

GDIT is teaming with AWS on Project Celerity, an AI-enabled platform for managing energy. The Army’s Advanced Research Lab has been heavily investing in how to deploy small, tactical “microgrids”—essentially energy generation and storage systems for environments where power and connectivity are absent. Those microgrids aren’t intended to simply provide power for soldiers on base but also power batteries for a growing fleet of eclectic and robotic vehicles and weapons. So predicting when and how drones or ground robots might require new batteries is a part of the challenge.

Shannon Judd, the director of global defense partners at AWS, said in an email that the military applications for the project are many. These include helping “teams conducting patrols or surveillance in remote areas, special operations forces who need to make decisions quickly without guaranteed access to full communications, and disaster or humanitarian missions,” on top of managing power or other pieces of infrastructure.

Brandon Bean, GDIT’s vice president for artificial intelligence and machine learning, said, “This is a proxy for contested logistics… The Baja Desert provides us with adverse terrain topography and weather; it also provides us a dynamic [operational tempo] so we can't pre-predict or plan anything.”

It shows an evolution of GDIT’s Defense Operations Grid-Mesh Accelerator, or DOGMA, a tool that fuses sensor data and sends it back to an operator under difficult conditions, such as enemy jamming, broken communications links, etc.

Since introducing DOGMA last August in the Pentagon’s T-REX drone warfare experiment, the company has developed three versions of it: one for fusing data, one for running autonomy, and one called WorldView, which Bean described as a “cognitive layer that provides a common operating picture.”

The race team will use electrically powered bikes, similar to the ones that special operations forces use in some missions. They’re quieter than motorbikes and their large batteries can also power sensors and communications gear.

“All the telemetry that's coming from the rider and from the motorcycle” will go to AWS servers, Bean said. “Then we're going to provide predictive analytics on when and where the rider needs to pit and where we need to replace the batteries.”

Other telemetry tools may eventually be added, like a rider-health tool designed for no-communication environments where standard fitness trackers don’t work. The company unveiled it along with DOGMA WorldView at the recent SOF Week event in Tampa, Florida.

“What we did was we built a round-loop workflow where we collected all this telemetry data off of these devices, [and] we're able to work and pull this data into our DOGMA WorldView and be able to do pattern of life on these individuals," Bean said. "So we could tell, based on the telemetry data on the phone, whether they've [encountered] elevated terrain or whether they stopped for periods of time. The next step of that is to actually tap into the microphone and the camera on the phone, so that we can identify if there's hostile control [of the] device.”

]]>

Thanks largely to robots, Ukraine is now talking about winning, not just surviving

Patrick Tucker — Tue, 02 Jun 2026 00:20:56 -0400

PRAGUE, Czech Republic–A small but growing number of European officials and analysts are saying what four years ago was unthinkable: Ukraine isn’t just surviving its grueling war with Russia, it is in some ways thriving and may even be on a path to victory.

This isn’t yet captured in headlines—for example, about last weekend’s barrage of Russian drones and missiles around Ukraine—but in the details, like how some 90 percent were intercepted.

Several long-term trends have shifted in Ukraine’s favor, and the core reason is its fierce focus on AI and robotics.

In the crucible of war, Ukraine has developed drones and ground robots that can hold territory—even take it back. Some are fully controlled by humans, like supply robots and medical-evacuation vehicles. But an increasing number are controlled in at least some aspects by dozens of AI products, from guidance packages on aerial drones to decision aids at the highest levels. Take the TFL-1 module, which can enable a one-way drone to function autonomously after a human has selected its target, reducing its susceptibility to jamming and other defenses. Its manufacturer, a Ukrainian company called The Fourth Law, says TFL-1 makes a drone four times more likely to hit its target.

Just as important as the tech are the new tactics. Given unusual latitude to experiment, Ukrainian fighters began to develop robot-forward infantry concepts, like combined-arms attacks by airborne and ground systems, “more than a year ago. Right now, we're massively starting to implement this,” said Davyd Aloian, deputy secretary of the National Security and Defence Council of Ukraine, the coordinating body on domestic and international security, in an interview.

Ukraine and its partners are also steaming ahead on new concepts for highly autonomous defenses against Russian drones, combining ISR sensors and AI to detect and identify enemy drones in less time and with more certainty.

“All of the systems are being linked with each other and with people” to create a distributed network with interceptor drones at various locations to be activated when needed, Aloian said. “One day we will have only like 10 guys who are just going to be responsible for approving interception. And it will automatically go direct to the target.”

The human operators will be dispersed as well. “Everything can be controlled from Kyiv, Lviv, from cities in other countries,” he said.

Ukraine’s advantages go beyond weapons and tactics. It is more willing than Russia—or even Kyiv’s Western backers—to rebuild its doctrine, acquisition, and resupply systems around autonomous warfare.

Countries that fail to follow suit risk disaster, one of Ukraine’s top dronemakers warned attendees at the GLOBSEC conference here.

“It's not what happened to Ukraine”—meaning Russia’s barrage of Shahed drones—that “should scare us in Europe,” said Swarmer CEO Serhii Kupriienko.

Instead, Kupriienko said, people should be scared by how quickly a middling military—in this case, Ukraine’s—developed the ability to inflict precise, devastating, and long-range damage.

“We are behind by literally 10 years or 20 years” in some defense-technology areas, such as satellite imagery, Kupriienko said, and yet his country has climbed a capability curve that just two years ago seemed insurmountable. So could others, he said.

“The answer is always AI solutions and integrating the AI into even the daily routine work within the bureaucracy,” he said.

Ukraine has also developed a defense industry that can keep up with the Russian threat. Its success is reflected not only on the battlefield, but in the growing number of foreign investors who see potential in defense products developed in and with Ukraine.

“We have evolved since 2022, the industry has and our defense has as well. Right now we are able to provide not only [large quantities of drone] assets but everything what is needed to build out the ecosystem,” including parts and production, training, modification, etc. Aloian said.

Strike drones FTW

Ukraine’s strike drones, more than any other factor, have helped counter a key Russian advantage: a large population of economically desperate young men and a comparative willingness to discount the cost of their deaths. Vladimir Putin has drawn hundreds of thousands into service with upfront bonuses and insurance benefits, which has provided numerical superiority on Ukrainian battlefields along with “considerable stimulus for the ailing Russian economy,” writes expatriate economist Vladislav Inozemtsev, who calls the system “deathonomics.”

But human waves are ineffective if drones kill soldiers faster than they can be replaced at the front—and that has become the case, analysts at the Institute for the Study of War wrote this week.

“Ukraine's successful mid-range and frontline drone strike campaigns are limiting Russia's ability to transport personnel to the frontline and to supply and sustain frontline positions,” they wrote.

Putin must now “convince an increasingly tired Russian populace not only to support a fifth year of war but also to accept involuntary mobilization for a war that has already cost Russia well over a million casualties.”

Ukraine’s deep-strike capabilities have changed the game in other ways as well. Oil infrastructure deep in Russian territory is no longer safe, giving Kyiv leverage over Moscow’s export revenues no matter what the White House does with sanctions relief. Even more humiliating, the drone threat forced Putin to hold his annual Victory Day parade this month without Soviet-esque ranks of tanks and missiles.

“Believe us. We were in the occupation of the Soviet Union for 50 years, and we know how important” the Victory Day parade is,” Estonian foreign minister Margus Tsahkna told the GLOBSEC audience. “For the first time, Putin was not able to wage this parade. This is the facade actually collapsing. And Putin is losing face among the Russian people, not only among us.”

“Putin thought that Ukraine was a question of five days. And, let's be frank, we, too, we said, ‘Five days, and then it's finished,’” said Xavier Bettel, Luxembourg's deputy prime minister. “In fact, the resilience of the Ukrainians was a big surprise for all of us.”

Changing fortunes

To understand how dramatically Ukraine’s prospects have changed, consider that in March, then-ODNI director Tulsi Gabbard, testified that the U.S. intelligence community believed that Russia had the “upper hand” in the conflict.

Now Ukrainian officials and other observers have begun to worry about a premature sense of victory among Ukraine’s foreign backers. Kyiv still depends on aid and imported weapons. Last week, Ukrainian President Volodymyr Zelenskyy said the government continues to be “very persistent” in its efforts to secure advanced Patriot missiles from the United States. “I believe [the U.S.] must act quicker,” he told reporters during a visit to Sweden.

Some European governments, however, are ever more eager to forge deeper ties with their continent’s new defense leader—not just for Ukraine’s sake but for their own.

“It means enlargement processes for the European Union, for NATO in the future,” Estonia’s Tsahkna said. It means security guarantees not only to Ukraine and for Ukraine, but the other way around, because actually Ukraine is the largest military power in Europe at the moment, and increasing as well its industrial base.”

As for the Ukrainian government, declaring victory will require more than the cessation of hostilities. The invading country must be left “much weaker,” so that it can’t re-arm as it did after its 2014 invasion of Crimea, Aloian said.

“If there's going to be a ceasefire, there will be very harsh conditions and difficult negotiations for the taking off of the sanctions, and when it will be,” he said. Otherwise, Russia will “renew all of those processes [of military buildup] before the full-scale invasion.”

“Right now, they're aiming like about 30 percent of their economy for the defense industry,” which is too much, he said.

Even the downfall of Putin, who has led Russia since the end of the 20th century, would be insufficient.

“The change of the regime shouldn’t just be only external. It should be also internal,” he said.

If it happens, much of the credit will go to the makers and operators of Ukraine’s drones.

]]>

The Navy wants next-generation munitions, so it’s spending millions on innovation hubs

Lauren C. Williams — Mon, 01 Jun 2026 16:41:19 -0400

Updated: June 2.

Replenishing munitions stockpiles doesn’t end with simply producing more of them. There are persistent challenges with mixing, manufacturing, integrating and discovering new energetic materials—the chemical compounds that make up explosives, propellants, and munitions. On Thursday, the Navy broke ground on a new facility to help.

The Maryland Energetics Innovation Hub is meant to furnish lab space where companies can test new tech, such as high-performance computers to run simulations. It is under construction an hour or so from the Pentagon by the Naval Surface Warfare Center in Indian Head, Md., and the American Center for Manufacturing & Innovation, or ACMI.

“This initiative ensures that NSWC Indian Head Division remains at the forefront of energetics innovation, scale-up, and production,” Capt. Stephen Duba, the warfare center’s commanding officer, said in a statement. “By bringing together government and industry partners in a collaborative environment, we can accelerate the development and fielding of critical capabilities that strengthen the Navy’s arsenal and the larger munitions industrial base.”

The Navy awarded ACMI $50 million to bring companies working in munitions and energetics closer to the service’s technical expertise. The group aims to raise another $200 million for the project.

The MEIH will focus on eight technical areas: developing new energetics materials, high-performance computing; non-destructive test and evaluation; integration with drones or unmanned systems; automating energetic processing and assembly; creating new manufacturing processes for propulsion systems and warheads; analyzing energetics obsolescence, and producing high-precision, high-throughput non-energetic components.

The plan is to finish the first two buildings of the larger facility within nine months, John Burer, ACMI’s founder, told Defense One.

The White House and Pentagon have made production of munitions—both exquisite and expendable—a clear priority for the near future. Earlier this year, the Pentagon stood up the Congressionally-mandated Joint Energetics Transition Office, which is charged with developing strategies for investment in and implementation of new and legacy materials needed for weapons and propulsion systems.

Defense tech hubs have already sprouted around the country, in Austin, Texas; Rhode Island’s Unity Park and Quonset Point; and the Louisiana coastline. MxD, a Pentagon manufacturing partner, has a 22,000-square-foot hub in Chicago.

The Maryland hub is the second facility the Navy and ACMI have embarked on. The first was the National Security Industrial Hub near NSWC’s installation in Crane, Indiana, which is focused on munitions and energetics. The Pentagon is spending $75 million to help erect the Indiana campus, which broke ground earlier this year.

The Maryland and Indiana facilities both are using private capital for “industrial facility build outs,” Burer said.

“The objective of the Maryland Energetics Innovation Hub is around process development and technology development around energetics, which would be developed there, but then scale and be relevant in many other places across the United States—qualifying new second sources of supply, which is a special thing that they have the ability to do at the nation's only government-owned, government-operated arsenal for the Navy, at Indian Head,” he said.

That munitions campus in Indiana is around 1,100 acres, while the Maryland hub would be a fraction of that.

“What makes this campus special is the ability to facilitate private-public partnerships between the tenants and Indian Head to make use of specialist capabilities that they have behind the fence,” such as mixing energetic materials, Burer said. “To build a solid rocket motor campus, for example, which is one of the specialties behind the gate at Indian Head, you need many, many hundreds of acres…That's what scaled production needs. But refining the processes at a pilot scale. It's smart to do that in a smaller footprint campus, in a collaborative way, which is what they're aiming to do here.”

The southern Maryland-based hub builds on an ecosystem of energetics companies working in the region, but could foster new growth and partnerships for companies designing military technology, said William Durant, CEO and president of the not-for-profit Energetics Technology Center, which will have space at the MEIH and help connect companies looking to work with the Navy.

“We want to see and help enable the companies that are coming in—that are best suited to meet any of those eight technical capability areas—[be] successful,” Durant said.

Over the next year, they plan to have a set of companies and a roadmap to execute solutions in those research and development areas.

“We want to see, in 16 to 18 months, who are the performers, what that's going to look like to meet the needs of the Navy…and then there are specific products and things that they're going to want that are needed today in the warfighter,” Durant said.

While there’s no fixed number, the aim is to have about ten companies join the hub, whether they take up long-term residence or cycle in and out, he said.

“Whatever is most important in supporting warfighter success. If that means a company comes in for six months, great! If that means that a company now needs to take up residency for five years, great! And now, does that mean we need to build another facility?”

Editor's note: This story has been updated to reflect the new funding goal ACMI officials set after they spoke with Defense One.

]]>