XC3 Weaponlight

Archive for the ‘AI / ML’ Category

« Older Entries

Army Researchers Modernize Breaching for Ground Platforms Through AI-Enabled Explosive Hazard Detection

Thursday, July 2nd, 2026

FORT BELVOIR, Va. (June 4, 2026) — To defeat adversaries’ explosive hazards on today’s battlefield, U.S. Army researchers are integrating the latest advances in artificial intelligence to deliver greater lethality and survivability to Soldiers.

With Soldiers facing increasingly sophisticated and complex threats, Army scientists and engineers are developing capabilities to enable persistent ground situational awareness for maximum force protection. The Army’s Command, Control, Communications, Computers, Cyber, Intelligence, Surveillance and Reconnaissance (C5ISR) Center leads the Ground-based Multi-Mission Payload project.

Breaching minefields has historically been one of the most dangerous tasks for troops. By automating the monotonous and fatiguing task of manual threat scanning, Soldiers can focus their attention on the broader tactical environment while easing the cognitive load. While unmanned aerial systems can cover wide areas, ground systems remain essential to detect threats aerial assets can’t see.

“Our S&T and technical expertise across core competencies including advanced sensing, intelligence, and command and control are delivering critical advantages for our Soldiers — situational awareness, enhanced operational speed, and safety,” said C5ISR Center Director Beth Ferry.

The GMMP proof-of-concept prototype includes a suite of hardware and AI-enabled software with advanced sensors, which have been outfitted onto a variety of ground vehicles and robotic platforms: a specially equipped military vehicle; a robot dog; and a Squad Multipurpose Equipment Transport, an unmanned, eight-wheeled heavy-duty robotic platform with instruments to complete multiple threat removal and complex mission sets, according to C5ISR Center physicist Kendall Johnson, the project’s technical lead.

An AI model detects, classifies, and reports explosive threats in real-time, integrating seamlessly into the Tactical Assault Kit ecosystem that populates a common operating picture for the entire team, both inside the vehicles and in the command post. Soldiers can identify hazards from a safe standoff distance, turning hours of manual scanning into a millisecond-fast automated process.

“The system incorporates a government-developed and -owned open AI architecture built by Army subject-matter experts,” Johnson said of the project’s plans for multi-algorithm support. “The Army can add the best algorithms from any source, at any time. The concept remains relevant into the future with the ability to incorporate new technologies as they emerge.”

C5ISR Center Countermine Ground to Ground Portfolio lead Dr. Amin Abbasi Baghbadorani said another project goal is transitioning from current counter-explosive systems that are often built with proprietary software and hardware while limited to a single purpose.

“GMMP is based on a modular concept to integrate commercial off-the-shelf hardware,” Abbasi Baghbadorani said. “Its open architecture is designed for rapid adaptation to new vehicles, sensors, and AI algorithms. The capabilities can be used with any platform and are easy to transition.”

Working with noncommissioned officers assigned to the Center is critical to providing Soldiers with the best tools for lethality and survivability, Johnson said.

“Feedback from NCOs has been incredible as we get feedback on-site,” Johnson said. “We’re able to make changes the same day and update the systems. It’s optimized the speed and pace of our project.”

Sgt. 1st Class Michael Havens, a C5ISR Center enlisted adviser, is working with the project’s scientists and engineers to bring his operational expertise as a network communication systems specialist into the technology development cycle.

“There’s an instant feedback loop,” Havens said. “What we do as enlisted Soldiers for C5ISR Center is they will give us their technology, show us how operate it, and run us through scenarios. We’ll tell them how to design the system to make it easier to use, more functional. Situational awareness is key. The more you have SA of the battlefield, the more you can devise a plan to execute, navigate, and negotiate.”

The GMMP team’s next steps are to mature the prototype into a cross-platform demonstrator with activities planned in additional climates and locations in the near future. It’s imperative the system performs across the wide range of conditions Soldiers face — extreme temperatures and humidity, sand, dust, foliage, snow, ice, and varying grass and soil types.

“The focus is adapting the system to more complex environments to prove its end-to-end capability,” Abbasi Baghbadorani said.

By Dan Lafontaine, DEVCOM C5ISR Center Public Affairs

Posted in AI / ML, Army, Breaching, Guest Post, Robotics | No Comments »

Space Force Integrates with Air Force in AI Sprint to Ensure Mission Dominance

Sunday, June 28th, 2026

LAS VEGAS, Nevada – To secure mission dominance in a future, contested environment, the Joint Force must make decisions faster than any adversary. This imperative was the driving force behind the Multi-Decision Advantage Sprint for Human-Machine Teaming, or MASH, a complex, two-week experiment recently hosted in Las Vegas.

Building on the successes of previous single-function Decision Advantage Sprints for Human-Machine Teaming experiments, the MASH marked a significant evolution by integrating an ensemble of artificial intelligence and automation software services from the first three DASH events. For the first time, U.S. Space Force Guardians joined Airmen to work side-by-side with software developers, evaluating how these disparate tools can effectively integrate to solve complex problems across the air, space, cyber, maritime, and ground domains.

“The Combined Joint All-Domain Command and Control Campaign Plan demands that we make better, timelier decisions,” said U.S. Air Force Col. John Ohlund, Advanced Battle Management System Cross-Functional Team director. “By incorporating AI into our battle management architecture, we are ensuring our operators can rapidly process vast amounts of data and deliver lethal effects faster than ever before.”

Conducted within a dedicated Shadow Operations Center-Nellis facility in Las Vegas, the MASH experiment set the stage for this strategic collaboration, led by the Department of the Air Force’s Advanced Battle Management System Cross-Functional Team. The experiment was executed in partnership with the Air Force Research Lab, U.S. Space Force, and the 805th Combat Training Squadron, also known as the ShOC-N, further reinforcing the collaborative effort required to deliver decisive combat power for the Joint Force. Furthermore, four allied nations observed the experiment, gaining insights into the U.S. approach to integrated architectures and setting the foundation for future interoperability.

Space Force Integration: A Critical Milestone

A defining feature of the multi-decision sprint was the active participation of Space Force Guardians. Moving beyond observational roles, Guardians were “in the seat,” directly influencing the development of battle management tools that encompass the space domain.

“Working with Air Force battle managers opened my eyes to how the air domain tackles these challenges. Their focus on tempo, synchronization, and rapid Courses of Action iteration mirrors what Space Force needs, especially when dealing with contested electromagnetic environments,” said U.S. Space Force 1st Lt. Abby Warner, 16th Electromagnetic Warfare Squadron deputy flight commander. “Turns out our decision-making headaches are similar across domains, and Transformational Model-based services adapt quickly to space ops.”

U.S. Air Force Lt. Col. Corey Ellsworth, ABMS Cross-Functional Team integration lead, agreed.

“There are parallels to decision advantage requirements between the air and space domains, especially during major combat operations where all domains are contested,”  Ellsworth said.

He noted that the next step for the DAF’s solution to battle management is to continue integrating with each service’s modernization approach to data and decision-making. The battle management software solutions tested at the MASH are “directly translatable” to Navy, Marine Corps, and Army partners, emphasizing that this collaboration is the next pivotal step in providing “combat multi-domain power” for the “Total Joint Force.”

U.S. Space Force Col. Teina Stallings-Lilly, ABMS Cross-Functional Team deputy director for space operations integration, emphasized the long-term impact of this integration.

“As the operations integrator between the services, my goal is to bridge the gap between our domains,” Stallings-Lilly said. “By having our Guardians in the seat for this experiment, they are seeing the direct applicability of these AI tools and, in turn, are providing the expertise needed to build a truly integrated DAF Battle Network.”

Stallings-Lilly explained that the DAF is moving beyond simple decision support systems to field capabilities that process information at machine speeds. This sprint, she noted, is fundamentally about building a human-machine team that ensures operators can think faster and stay decisively ahead of any adversary.

The need for deep, cross-service integration extends far beyond the air and space domains, shaping the future of command and control.

“The reason we challenge the software to solve multi-domain problems is because that’s the reality of the future fight,” said Ohlund. “An Air Force air battle manager doesn’t have the authority to execute a space or cyber effect, but like any good staff officer, it’s their job to prepare the information and package the options for the general. We want the computers to do that work, to ruminate over every possible multi-domain effect; that way we can present the highest quality menu of decisions to the right commander, faster than ever before.”

WARTECH: Co-Creation for Rapid Fielding

This deep integration of multi-domain warfighters into the development process is a key component of the larger  AFRL process known as WARTECH, which brings together warfighters, technologists, planners, and acquisition personnel to collectively develop operational concepts motivated by future force design and enabled by high-payoff science and technology.

“The DASH to MASH series is really a textbook example of what WARTECH is intended to accomplish and right in line with the Command, Control, Communications, and Battle Management strategy for agile, rapid, and iterative fielding of software solutions to support immediate warfighter needs and long-term force modernization,” said Jeffrey Palumbo, AFRL C3BM Capability Area lead. “This approach of user-producer co-creation allows for proof of concept, energizes the industrial base, allows for early operator feedback to shape development, and sets us up to deliver chunks of decision advantage capability to the warfighter in a rapid and repeatable cycle.”

The MASH Ensemble: Perceive Actionable Entity, Match Effector, and Generate Battle COAs

The experiment challenged six industry software development teams and the ShOC-N’s own military software development team to build tools that address three core decision functions derived from the DAF’s Transformational Model:

PAE: Recommending what actions can be taken against a target.

Match Effector: Given a list of possible effects, ranking a capability or a set of capabilities best suited for the given effect, and repeating for each of the other provided effects.

Generate Battle COAs: Given a list of matched effect-effector pairs, adding the additional capabilities throughout the execution window needed to support the principal match, and repeating for each of the next ranked pair.

A major breakthrough of the event was the successful integration of these disparate vendor tools.

“AFRL has done incredible work building an orchestrator that ensures these different companies can exchange data, ontologies, and metadata seamlessly,” Ohlund said. “We are proving that a true plug-and-play, modular approach not only works, but it fosters continuous competition and allows the government to select the best-of-breed software services as they mature.”

The Warfighter as Expert Evaluator

Throughout the sprint, the Airmen and Guardians were tasked not just as operators, but as expert evaluators. Their mission was to stress-test the AI’s decision logic, identifying limitations and providing immediate feedback to the developers sitting directly behind them.

“This is a true co-creation environment where software developers work directly with warfighters to ensure the tools meet their exact needs,” said Elizabeth Frost, AFRL MASH lead. “The teams are eager for feedback and implemented changes rapidly. This collaborative effort paid off during the second week of the sprint, as we saw a remarkable increase in the volume and quality of courses of action submitted.”

The operational impact of this co-creation was immediate and undeniable for the tactical operators.

“A week ago, it took my team and me 50 minutes to an hour to get one tasking done. With the help of the tool, we were able to get five or six taskings done,” said U.S. Air Force Capt. Adam Sochia, 552nd Operations Support Squadron ABM. “Basically, in the amount of time that we can do one tasking, this tool gives us the data and accurate options to complete five or more additional taskings.

Delivering a Lethal, Integrated Future

The event also featured the ShOC-N’s military software development team, who built their own solutions alongside industry. According to Carlos Dye, the ShOC-N MASH software development team lead, the military developers focused on applying their direct operational experience to the coding process. Their approach ensured that the machine took the brunt of the data processing, while the human operator remained firmly in control of the final tactical decisions.

This unique environment, which physically co-located military operators, Airmen developers, and industry partners, was critical to the event’s success.

“The synergy we are seeing here… is what has been lacking in previous attempts to accelerate delivery of warfighter capability,” said Lt. Col. Wesley Schultz, 805th CTS/ShOC-N commander. “Our mission at the ShOC-N is to remove barriers to creative problem-solving, allowing us to turn innovative concepts like human-machine teaming into tangible, lethal capabilities at speed.”

A key factor in enabling that speed and synergy was the underlying technical framework. Elizabeth Frost, the AFRL MASH lead, noted that by establishing a common application programming interface and architecture, the team was able to provide a unified user interface. This meant that regardless of which vendor’s software was running in the background, the experience remained consistent and intuitive for the warfighter, proving that integrated tools deliver a far better outcome than isolated solutions.

Ultimately, the MASH experiment provided an actionable blueprint for the future of multi-domain operations. The event validated the DAF’s Transformational Model, proving that when battle management is broken down into specific decision functions with a common integration framework, machines can process data at a speed unmatched by humans.

Ohlund concluded, “By demonstrating that diverse, AI-enabled tools can integrate effectively within this model to accelerate the kill chain, the DAF has taken a critical step toward securing decision advantage for the Joint Force.”

Deb Henley

505th Command and Control Wing

Public Affairs

Posted in AI / ML, Air Force, C5ISR, Guest Post, Space | No Comments »

Army Innovators Automate Path to Zero Trust with Artificial Intelligence

Sunday, June 21st, 2026

The Communications-Electronics Command Army Software and Innovation Center in partnership with the Warfighting Acquisition University and the C5ISR Center, has developed an artificial intelligence tool that significantly accelerates the Army’s transition to a Zero Trust cybersecurity framework. This supports the Department of War’s mandate to achieve Target Level Zero Trust by Fiscal Year 2027.

Zero Trust is a cybersecurity strategy based on the idea that networks are always at risk. Instead of trusting devices within the network, it requires every user and device to be authenticated and authorized before accessing data. Rolling out this approach across all Army systems is a big challenge, but it is necessary for operational readiness.

“Many organizations know they need Zero Trust, but I believe they’re overwhelmed by not knowing where to begin,” said Farhat Shah, a cybersecurity subject matter expert with CECOM ASIC.

Shah discussed this challenge and its corresponding solution during the Warfighting Acquisition University event, “Operationalizing Zero Trust – Leveraging Risk Management Framework and Artificial Intelligence,” held on May 13.

During the presentation, Shah explained that the team started with a five-month project to crosswalk the 91 Zero Trust activities to the thousands of Control Correlation Identifiers in the Army’s Risk Management Framework. This method aligns efforts down to the CCI level, which Shah calls “critical, because CCIs are actionable, testable elements that we use during our RMF assessments.” This helps system owners use their existing compliance work to check their Zero Trust status.

“Our goal is to reduce duplication of effort,” Shah said. “We want to save time and resources by leveraging existing work, and most importantly, reducing risk in a manageable and sustainable way. It is about aligning strategy, governance, and technology.”

The core innovation is “AI Flow,” an AI environment developed by CECOM ASIC. This tool processes a system’s RMF test results and automates the analysis to generate a Zero Trust baseline profile. In a pilot assessment of the Army Food Management Information System, the AI completed the review in about five minutes, compared to a week for a human expert. The assessment found that the AI was 89% accurate.

This system works with two agents. The first agent checks for compliance. If a system is noncompliant, the second agent investigates further, identifies specific gaps, and provides clear guidance, including references to the relevant policies and required documents. This transforms a simple compliance check into a step-by-step engineering process.

With the FY 2027 deadline approaching, CECOM ASIC seeks to partner with additional system owners to expand the tool’s capabilities and help them quickly assess their Zero Trust posture.

“As we continue to refine this process, we’re not just improving the tool; we are shaping a repeatable and scalable approach to support Zero Trust adoption across the enterprise,” said Shah. “If you are interested in advancing Zero Trust automation or want to see how this approach can benefit your organization, I invite you to partner with CECOM ASIC to evaluate and refine this process.”

This integration of existing frameworks and artificial intelligence provides a scalable, data-driven roadmap, that enables the Army to secure its systems and protect its data amid evolving digital threats.

For those who missed the May 13 session, the presentation and materials are available online. Personnel in the Defense Industrial Base, academia, and IT or cyber communities can access the recording and slides to learn how the Army is advancing cybersecurity. Watch the full presentation here: events/operationalizing-zero-trust-leveraging-rmf-and-ai

For inquiries or collaboration, contact ASIC Cybersecurity & Electromagnetic Warfare Directorate at Usarmy.apg.asic.mbx.zero-trust@army.mil.

By SCOTT HOCHENBERG

Posted in AI / ML, Army, Guest Post | No Comments »

Gallatin AI Awarded Contract by US Army’s III Armored Corps

Friday, June 12th, 2026

WASHINGTON, June 11, 2026 — Gallatin AI today announced it has been awarded a contract by the US Army’s III Armored Corps to deploy and refine Navigator, its AI-native logistics decision support suite, in direct support of III Corps exercises and operational planning over the next 18 months.

The award addresses a capability gap at the corps echelon where Army ground combat units lack tools that enable sustainment decision support at the operational level of war. Under the agreement, Gallatin will deliver a tailored Logistics Common Operating Picture (LOGCOP); predictive consumption algorithms across all classes of supply; AI-assisted planning tools for the Military Decision Making Process (MDMP); and integration with Next Generation Constructive (NGC) simulation platforms. The effort maps to two Department of War critical technology areas: Applied Artificial Intelligence and Contested Logistics Technologies.

“Planning and executing Sustainment at the operational level is fundamentally different from what happens at the tactical or strategic echelon,” said Woody Glier, CEO of Gallatin AI. “A corps commander and staff must forecast what tens of thousands of Soldiers will require over months of sustained conflict while placing those requests against theater sustainment commands and depots that may be a continent away. That requires a decision support capability built for operational timescales and operational complexity.”

“At the corps level, the hard part is reconciling demand against supply,” said Brian Ballard, Chief Product Officer at Gallatin AI. “A staff is pulling a noisy, continuous demand signal from formations that may be dispersed across more than one theater, and it has to turn those disparate pieces of information into supportable courses of action (COAs) for validation with an Corps Sustainment Command (CSC) before anyone commits resources. Navigator surfaces the right data at the right time and lets corps and CSC planners develop and stress-test those COAs together.”

Navigator is already deployed with multiple military units. Work under the III Corps Other Transaction Agreement (OTA) will be validated through iterative delivery across multiple live exercise events.

About Gallatin

Founded in 2024 and backed by 8VC and leading defense and technology investors, Gallatin AI develops capabilities that make the Joint Logistics Enterprise predictive, visible, and accountable, from the point of production to the point of need. Headquartered in El Segundo, CA with offices in Washington, DC and Austin, TX, Gallatin’s flagship platform, Navigator, is deployed with multiple military units enabling precision sustainment at the speed of relevance.

For more information, visit www.gallatin.ai

Posted in AI / ML, Army, Contracts, Press Release | Comments Off on Gallatin AI Awarded Contract by US Army’s III Armored Corps

SPARC AI Expands Overwatch Targeting Capability with Image Recognition and Successful 43km Target Acquisition Test

Wednesday, June 10th, 2026

9 June, 2026: SPARC AI Inc. (the “Company”) (CSE: SPAI) (OTCQB: SPAIF) (Frankfurt: 5OV0) a defence technology company building Overwatch, the GPS denied navigation and target acquisition software platform for drones and autonomous systems, today announced the successful completion of a 43km long-range target acquisition test conducted over open water in Port Phillip Bay, Victoria, Australia. The target recording was done at a drone height of 115m above ground level.

The 43km demonstrated span is comparable to, and in some measurements exceeds, the narrowest width of the Strait of Hormuz, one of the world’s most strategically significant maritime chokepoints. The comparison illustrates the scale of contested, GPS-denied maritime environments in which the capability is designed to operate.

SPARC AI is also pleased to announce it has integrated image recognition into the SPARC AI drone controller application, adding further capability to its targeting solution. Overwatch brings together targets recorded by multiple drones across different manufacturers and different locations onto a single operating map, where operators can classify and track targets, collaborate, and plan missions in one shared picture. With image recognition now overlaid onto that picture, operators gain richer intelligence and can respond more rapidly across teams.

Capabilities of this kind have historically been locked inside expensive, proprietary drone platforms. By delivering them as software across any manufacturer’s hardware, the Company believes Overwatch meaningfully expands its addressable market and positions the platform as a premium software layer rather than a single-aircraft feature.

Looking ahead, the next phase of Overwatch’s development will introduce the ability to deploy multiple drones directly from the platform. The company is developing teaming and swarm capability that it believes will be unique to Overwatch with the ability to deploy and coordinate drones from different manufacturers, operating from different locations, simultaneously and to do so in a GPS-denied environment.

SPARC AI intends to make these capabilities available to its partners in Dubai, Ukraine and the United States with the next software update.

Posted in AI / ML, Press Release | No Comments »

Leveraging Artificial Intelligence to Ensure Secure Army Communications

Sunday, June 7th, 2026

FORT HUACHUCA, Ariz. — In today’s rapidly evolving digital landscape, the convergence of large datasets and AI is transforming the way the Army secures and operates its networks. The U.S. Army Information Systems Engineering Command, a major subordinate command of the Communications-Electronics Command, stands at the forefront of this transformation by harnessing the power of AI to enable advanced data analytics and drive Army operational readiness and resilience across the force.

As Army networks expand and the volume of data grows, the attack surface for our adversaries also expands. Recognizing this, USAISEC is leading an effort to leverage large datasets with AI-driven insights to enhance operations and harden critical infrastructure for secure Army communications.

Supporting communications security

USAISEC is the home of the Army’s COMSEC experts. Tasked with supporting COMSEC across the Army, USAISEC ensures the proper safeguarding and handling of cryptographic materiel for hundreds of Army units worldwide. These units are engaged in missions ranging from training and humanitarian assistance to real-world military operations around the globe that require 24-hour support.

The USAISEC Communications Security Directorate provides critical support across three primary focus areas:

COMSEC accounts and keys: CSD operates the Tier 1 system that generates and distributes cryptographic keys to Key Management Infrastructure workstations. To maintain strict accountability, personnel serve as the Service Authority and Central Office of Record, providing oversight and assistance for all Army COMSEC account management.

COMSEC operations: Warfighters require continuous support to maintain secure networks in contested environments. To assist account managers with daily operations, CSD strategically positions regional COMSEC Information Security Representatives at geographic locations worldwide. Additionally, personnel staff dedicated help desks to guide soldiers through the complex operation of Key Management Infrastructure workstations, cryptographic hardware, and associated software.

COMSEC policies and compliance: Security is effective only when standards are strictly enforced. To enforce policies, CSD acts as a COMSEC Incident Monitoring Activity, actively assessing and mitigating reported security incidents across the Army. Additionally, personnel develop modernized procedures and conduct thorough audits of Army COMSEC accounts to guarantee absolute compliance with stringent national security policies.

Using AI to drive predictive analytics

This immense responsibility generates vast amounts of data, which historically required manual review and management. Today, USAISEC is leveraging AI-driven data analytics to assess these large datasets, enabling predictive analytics that allow for proactive engagement with Army units before challenges arise. This not only heightens Army readiness but also empowers commanders and decision-makers with data-driven insights to better manage COMSEC programs within their force structures.

One of the most significant advancements in this modernization effort is the development of the Communications Security Operations Center. Rather than altering the current mission, the CSOC strengthens and streamlines core COMSEC services by centralizing operations, improving readiness visibility and enabling a more integrated, data-driven support model. For the first time ever, data previously available only to CSD personnel is now available to commanders at all levels.

When fully implemented, the CSOC will support improved Army readiness by allowing detailed analysis of COMSEC account health, compliance and risk, while enabling proactive mitigation and forecasting of cryptographic materiel requirements. Powered by continuous data integration from across CSD’s mission areas, these enhancements ensure Army networks remain resilient and secure.

USAISEC is deploying AI to proactively monitor the Army’s COMSEC posture. By leveraging data-driven insights and establishing a centralized CSOC, USAISEC is shifting COMSEC management from a reactive to a predictive model. These advancements enable the Army to maintain continuous, secure communications for our warfighters worldwide. The resulting Army resilience ensures Army networks remain ready for complex operations across all domains.

By Steven Downer and Sandra Rosario, USAISEC

Posted in AI / ML, Guest Post | No Comments »

Merlin Successfully Completes Critical Design Review for C-130J Autonomy Program with USSOCOM

Friday, June 5th, 2026

CDR approval marks the completion of the final design configuration for the C-130J, advancing the program to the aircraft integration phase

BOSTON, June 04, 2026 — Merlin, Inc. (NASDAQ: MRLN), an aerospace and defense technology company building the operating system of record for autonomous flight, today announced the successful completion of the Critical Design Review (CDR) for its C-130J autonomy program with the U.S. Special Operations Command (USSOCOM). Completion of the CDR is a milestone in the program’s maturation, validating the system’s design readiness and advancing Merlin from design development into aircraft integration activities.

This milestone positions the program to enter a structured formal test campaign, including aircraft-level testing, reflecting a disciplined systems engineering progression from design through verification.

The CDR is part of Merlin’s previously awarded IDIQ contract under its C-130J autonomy program with USSOCOM. The milestone supports a demonstration of Merlin’s C-130 autonomous capability development, showcasing the potential reduction in crew workload through all phases of flight. Under this contract, Merlin is rapidly advancing its AI-powered autonomy stack onboard the C-130J, with potential pathways for expansion across other Department of War or commercial aviation platforms.

“Completing the Critical Design Review validates the architecture we’ve built for safe, scalable autonomy on large aircraft like the C-130J,” said Matt George, CEO and founder of Merlin. “We are grateful for USSOCOM’s partnership and guidance in this effort as CDR is another important step in demonstrating that our system meets the rigorous standards required for our USG customers’ real-world deployment. As we move into integration, ground testing, and eventually flight demonstrations, we’re focused on proving autonomy from takeoff to touchdown is one of the most effective ways to improve operations and safety for US warfighters.”

Posted in AI / ML, Air Force, Aviation, Press Release, SOF Week | No Comments »

Dear Industry — Stop The Slop: Part Deux

Tuesday, June 2nd, 2026

A few weeks ago I put out a plea to industry to not use Artificial Intelligence Large Language Models to create press releases. Something happened the following week which really lit a fuze under me. Not long after, something happened which made me double down.

I was attending SOF Week and happened across a company specializing in AI. Why they use it in an entirely different way than LLMs, I shared the story of running across a PR so egregious that I created my post. A little laugh was had by all.

Not long after, I received that very company’s PR for the show. It was even worse than the one that had set my jaw the week earlier.

I headed back to the booth and as I walked up the head of Marketing knew exactly why I had returned. I showed them the PR and what the back end looked like.

It had been written by their Public Relations Agency and the company I was speaking with contacted that agency to inquire about their product. I too emailed the representative who had sent me the PR to inquire as to whether AI had been used and included a screenshot of the HTML version of the PR in question which was an absolute hot mess.

That evening I received this reply.

When I read the reply I let out a loud curse. Those of you who have to deal with slop know why.

You see that extra long dash in the email? That’s called an “emdash” and nobody on this planet uses that damnable punctuation mark except for AI. For some reason it’s so insidiously embedded in AI that if you tell an LLM not to use it, it will go back to it very quickly. You see an emdash, it’s 99.9% AI derived slop.

To summarize, I send out an inquiry as to whether or not a document was written using an LLM and the reply I received denying its use was either written using AI, or was a serious prank. Pretty bold move.

The next morning, I returned to the company the PR had been written about to discuss the issue with their head of marketing. Initially, they did not know the significance of the use of the emdash. They then made me privy to an email written in response to their inquiry with their PR agency. That email disclosed that they had indeed used AI to assist in creating the PR. This was at odds with the assertion I had received from my POC.

I don’t like being lied to. I am in my late 50s and the father of six children, I know when someone is BSing me. At that moment, I made a decision to no longer accept any PR submissions from that particular representative, no matter where they went to work, and from that PR agency in general.

I’m not going to throw them under the bus publicly, but hopefully they read this and are ashamed knowing that I’ve got their number.

As for other agencies out there, it’s OK to use an LLM to conduct research or to assist with grammar and readability. But don’t let it do all of your work for you.

Customers read the slop and know that it’s junk. AI agents scour the web, searching for slop and let search engines dough to avoid recommending websites that are full of slop.

(more…)

Posted in Admin, AI / ML | 9 Comments »

« Older Entries