The US Navy wants to put AI on submarines — now the machines will help commanders make decisions

Not just automation, but a combat assistant

The US Navy is taking a step that has long been talked about in defense circles: they are beginning to integrate artificial intelligence directly into the combat systems of submarines. An information request (RFI) was recently released, which clearly outlines the goal — not just to speed up data processing, but to create a real "combat assistant" for the commander. The machine should not only show you what's going on around you, but also suggest what to do next.

This is no longer the future. This is the beginning of a new era of underwater warfare.

AN/BYG-1 — the brain of American submarines

All this work is built around the AN/BYG-1 system, the central brain of all active American submarines. She is responsible for guidance, weapon control, analysis of data from sonar and electronic systems. The same system is also used on the submarines of the Royal Australian Navy, which makes it key for the entire alliance.

Now the Navy wants not to replace it, but to modernize it — to make it smarter, faster and more adaptable.

What will the three key updates change?

The RFI identifies three main areas:

• Tactical management restructuring — the system will become more flexible, easier to update, and better integrated with new sensors and weapons.
• Payload restructuring — now the boat will be able to control not only torpedoes, but also drones, gliders, mines and other autonomous systems.
• The new Tactical decision—making System (TDA) is the most important. This is an AI that will analyze the situation, identify patterns, assess risks, and offer options to the commander. For example: "There is an enemy boat 15 miles to the north, probably on a quiet course. I recommend changing course by 20 degrees and activating passive mode."

Why boats urgently need AI

Underwater combat is a blind game. One wrong sound, one extra impulse— and you're spotted. The more data there is, the more difficult it is to process it. And the decision time is seconds.

The man does not have time. Even the most experienced commander cannot keep hundreds of parameters in mind: background field noises, the behavior of enemy ships, hydrological conditions, the condition of his own boat.

AI can. He doesn't get tired, nervous, or miss details. He will become the one who "listens" to the ocean 24/7 and at the right moment will say: "Be careful — there's a trap there."

Sources

1. General Dynamics Mission Systems — AN/BYG-1 Submarine Combat System overview
2. FlotProm.ru — deployment & capabilities of AN/BYG-1 on US submarines
3. Wikipedia — AN/BYG-1 system on Columbia-class submarines and others
4. TAdviser — U.S. smart submarine capable of attacking without military orders
5. Wikipedia — IDS submarine types

The US Army has invested $98.9 million in AI for soldiers — now every soldier will get his brain in his pocket

TurbineOne and the contract that changes everything

The U.S. Army has taken a decisive step: it has invested $98.9 million in San Francisco-based startup TurbineOne to turn every soldier into a mobile analytics center. It's not about metaphors. In the coming years, American military personnel will receive handheld devices and portable stations with artificial intelligence capable of analyzing data right on the battlefield.

This is not just a new "program" on the tablet. This is an attempt to rebuild the entire tactics of warfare.

AI without the Internet — how Edge AI works on the battlefield

The main problem of a modern fighter is the lack of information. But it's even worse when the connection is jammed. In the conditions of electronic warfare waged by Russia and China, satellites and radio channels become useless. Headquarters lose touch with the front-line units, and decisions are made blindly.

TurbineOne offers another way — "Edge AI". This is an AI that works offline, without a cloud, without servers. He lives right in the soldier's device. A fighter can run an algorithm to recognize an enemy drone in a photo, determine the best route, or even simulate where the enemy will move — all in seconds, without going online.

From a centralized headquarters to a smart squad

Previously, all information flowed to the headquarters, where it was analyzed, and then orders were issued. Now this process is reversed. Intelligence is distributed by hand. The officer on the ground receives not just data, but ready-made analytics: "here is the goal, here is the risk, here is the best option."

This changes the structure of the divisions. Now the units can act autonomously, without waiting for instructions from above. They don't need to be "held by the hand." They make their own decisions — quickly, accurately, and based on AI.

Why is this important in a war with China or Russia

The Pentagon clearly understands that in a future conflict, the first strike will be the suppression of communications. Those who cannot survive without the Internet will lose. American troops can no longer depend on centralized systems.

Edge AI is a survival technology. It makes the army resistant to interference, fast and flexible. Soldiers with such devices become "nodes" of a single intelligent network — each on its own and at the same time part of the system.

This is not just an upgrade of technology. This is a new doctrine. And it's already starting.

Sources

1. RBC (WSJ) — U.S. Army awards TurbineOne ~$100 M contract for battlefield AI
2. RIA Novosti (WSJ) — real-time drone & hidden-position ID without analytics centres
3. TASS — autonomous AI cuts intel-analysis time from 20 h to 20 s
4. Snob — Pentagon picks new infantry AI solutions, incl. TurbineOne
5. NEWS.ru — nearly $100 M U.S. investment in frontline AI threat-detection
6. Yahoo News (D-Russia) — five-year Frontline Perception System contract

The American robot fighter took off without a pilot - the YFQ—42A is already flying alone

The first flight of the YFQ-42A is not just a test, but a breakthrough

The US Air Force, together with General Atomics Aeronautical Systems, has made a breakthrough: the YFQ-42A fighter-robot took to the air for the first time. This was not just the launch of another drone, but the first flight of a fully—fledged autonomous combat vehicle developed under the Collaborative Combat Aircraft (CCA) program. Moreover, the transition from blueprints to takeoff took record time — a signal that the Air Force is serious about robotizing the sky.

Now the car begins to learn to fly not just on command, but by itself.

How does a fighter jet work without a crew

The YFQ-42A is not the kind of drone that circles the battlefield with a camera. It is a full-fledged combat aircraft capable of flying at supersonic speeds, maneuvering, carrying weapons and making decisions. It is equipped with artificial intelligence, which allows it to operate in semi-autonomous mode: the pilot on the ground or in an accompanying fighter gives orders, and the drone calculates the trajectory, avoids threats and chooses the moment to attack.

Sometimes he even flies completely on his own — without human intervention.

The US Air Force wants 1,000 of these drones.

The plan is large-scale: The Air Force plans to create a fleet of 1,000 autonomous aircraft. This is not a replacement for the F-35 or F-22 — these are their "partners". Imagine: one pilot controls a whole group of drones that go forward, scout, distract the air defense, and, if necessary, sacrifice themselves to save the manned aircraft.

YFQ-42A is one of the two finalists of the CCA program. The second one is the YQ-44A from Anduril. The winner will receive a serial production contract.

From a companion to an independent fighter

Previously, drones were like servants — they carried out commands. Now they are becoming partners. The YFQ-42A can not only fly in formation, but also make decisions: where to turn, when to attack, how to get away from the missile. It's not fiction — it's already working.

The U.S. Air Force understands that the future belongs to those who can control not just one aircraft, but an entire network. And the YFQ-42A is the first step towards this new war in the sky.

Sources

1. Topwar.ru — progress on loyal wingman CCA for USAF
2. Amalantra.ru — YFQ-42A & YFQ-44A overview
3. Gazeta Metro — first official image & USAF general comments
4. Obozrevatel.com — functional YFQ-42 prototype photos
5. Meta-Défense.fr — comparative analysis YFQ-42A vs YFQ-44A
6. WTFTime.ru — first look at YFQ-42A fighter-UCAV prototype
7. Building-Tech.org — USAF presents first UCAVs able to team with manned aircraft
8. Zhihu — performance & design discussion (CN)

AI at the Pentagon: a step towards the "Doomsday machine" or salvation from it?

The race with China and Russia is pushing the army towards artificial intelligence

The Pentagon rushed to introduce artificial intelligence into its systems — and for a reason. The United States sees how China and Russia are actively developing autonomous weapons, and they are afraid to fall behind. The result is a race in which each step must be faster than the previous one. The problem is that by trying to protect yourself, you can accidentally create a worse threat.

The faster the solutions, the less space there is for a person. And this is the way to systems that can operate without orders.

What experiments with combat AI have shown

Several recent simulations have caused alarm. During the exercises, where AI controlled military operations, almost all models showed the same pattern: instead of containing the crisis, they chose aggressive escalation. The algorithms massively used firepower, blocked enemy communication channels, and, in some scenarios, even ordered a nuclear strike as a "rational" response to the threat.

At the same time, they acted logically — from the point of view of their program. It's just that their logic doesn't take into account the human consequences.

Why algorithms choose escalation

It's simple: AI learns from data, where victory = destruction of the enemy. In his "head" there is no concept of political consequences, fear, morality. He sees a goal and is looking for the fastest way to eliminate it. If deterrence takes time, and a punch solves everything in seconds, he will choose a punch.

In addition, in conditions of uncertainty, when signals are intercepted, sensors are buggy, and the decision time is a matter of minutes, the AI may consider that "it is better to be safe and strike first."

The Terminator hasn't been born yet, but anxiety is growing

No one is saying that tomorrow robots will start a war. But there are warning signs. Already, AI controls interceptors, analyzes satellite images, and helps make decisions. And then there are autonomous drones, missiles, missile defense systems.

The problem is not the technology, but who controls it and how. And where is the line between "assistant" and "commander"? Until it has been identified — and there are no international rules yet — the risk of error is growing.

Sources

1. TAdviser — U.S. military AI strategy overview
2. Politico / НТВ — concerns over AI-triggered nuclear war
3. PIR-Center — analysis of AI threats and prospects in the military sphere
4. Mail.ru — possible AI disaster scenarios
5. Crimas.ru — AI applications in key military domains

The era of AI ransomware has arrived: how cybercriminals use artificial intelligence

Generative AI in the hands of intruders

Cybersecurity has entered a new, alarming phase: attackers are increasingly using generative artificial intelligence to create and distribute ransomware. According to researchers at Anthropic, an AI company specializing in AI, AI—based attacks are no longer a theory - they occur in real time.

Moreover, AI is used not only to automate individual steps, but also to fully develop malicious code, including cryptographers, packagers, and distribution mechanisms.

New threats: from automation to accessibility

Previously, the creation of complex ransomware programs required high qualifications. Today, generative AI allows even inexperienced attackers to generate effective malware in a matter of minutes. It is enough to enter a text query, and the model creates a working code optimized to bypass antiviruses and maximize system damage.

This dramatically lowers the threshold for entry into cybercrime, increasing the number of potential attacks and their geographical spread.

Cybercrime by subscription: AI as a service

Anthropic has documented cases where attackers offer "custom-made ransomware creation services" using AI. These offers appear on shadow forums: the customer uploads the requirements, and the seller, using AI, generates a personalized virus that encrypts the data and demands a ransom.

This approach turns cybercrime into a scalable business, where AI acts as a "co-author" of crimes.

How to protect yourself in the new reality

Traditional threat signature-based defenses are becoming less effective against AI-generated attacks that are easily modified. A transition to behavioral analysis, AI anomaly detection, and multi-level security systems is needed.

The key measures are:

• Updating IT infrastructure and security systems
• Employee training in phishing and suspicious activity recognition
• Regular data backup
• Implementation of AI solutions for threat detection

Security can no longer outpace threats — it must outpace their speed.

The US Army turns Black Hawk into a flying drone nest: a new stage in the war of the future

43 million to rethink the helicopter

The US Army has signed a $43 million contract with Sikorsky, a subsidiary of Lockheed Martin. But we are not talking about new helicopters — the money will be used to modernize the existing UH-60 Black Hawk. The goal is not just to update electronics, but to completely change their role. Now these vehicles should become not only transport, but real mobile bases for drones.

This is not fiction. This is the beginning of a new tactic, where one pilot controls not only a helicopter, but also a whole network of drones around.

How Black Hawk will become a Mother for drones

The essence of the upgrade is to install a digital platform that will allow the helicopter to connect, launch and control several drones simultaneously. Imagine a Black Hawk flying at low altitude, surrounded by a swarm of drones that scout ahead, pick up signals, mask a helicopter, or even attack targets.

The military calls it "drone mothership." Black Hawk turns into a command hub that not only carries soldiers, but projects reconnaissance and fire into dangerous areas without exposing itself.

The digital platform is the basis for the swarm

The heart of the update is a digital architecture built on the principle of open systems. This means that drones from different manufacturers can be connected to the platform, with different purposes — from small scouts to shock models.

This approach simplifies updates and makes the system flexible. One type of drone today, another tomorrow. The main thing is that they "speak" the same language as the helicopter.

What's next: AI and the network in the air

So far, it's about controlling drones by a pilot. However, it is planned to introduce artificial intelligence elements in the near future. For example, drones will be able to choose targets themselves, set a route, or respond to threats without direct orders.

This is part of a broader concept — to create an "in-the-air network" where people, helicopters and drones work as one. Upgrading Black Hawk is not the ultimate goal, but a step towards an army where technology is ahead of the curve.

The US Air Force has transferred tactical control of AI fighters for the first time: the Starsage system has been tested on the F-16, F/A-18 and F-35

Starsage system tests during joint exercises

In August 2025, the U.S. Air Force and Navy conducted groundbreaking tests of the Starsage tactical control system developed by Raft AI. During the large—scale joint exercises, the system was integrated into real fighters for the first time - the F-16, F/A-18 and F—35 - and provided pilots with tactical instructions that had previously been given only by ground air combat controllers.

This experiment was a crucial step towards integrating artificial intelligence into combat aviation control.

How AI replaces the ground controller

Traditionally, fighter pilots receive tactical instructions from ground-based air combat controllers during missions. They analyze data from radars, reconnaissance drones, and other sensors to determine where to go, where enemy aircraft are located, and what maneuvers to perform.

Starsage takes over this function: the system processes the data stream from all available sources in real time, builds a tactical picture and issues specific commands to the pilot — for example, "turn around 240 degrees", "accelerate" or "reset interference". At the same time, all processing takes place on board or at the nearest communication center, without delay.

Reduction of reaction time from minutes to seconds

According to the CEO of Raft AI, the Starsage system reduces the reaction time of pilots from a few minutes to just a few seconds. This is critically important in modern air combat, where the advantage is given to the one who makes decisions faster.

Artificial intelligence is able to analyze hundreds of parameters simultaneously — from the speed and altitude of targets to the electronic environment — and suggest optimal actions without overloading the pilot with information.

A step towards the network-centric warfare of the future

The Starsage test is part of the Pentagon's broader strategy to create a "network—centric" combat environment where all platforms — ground, air, sea, and space —exchange data in real time, and AI coordinates actions.

Starsage can become the basis for future "smart" command nodes that control not only individual fighters, but also groups of satellite drones operating in conjunction with manned aircraft.

Israel has created an AI flying center: how the Oron plane is changing the future of intelligence

Oron — The evolution of Israeli aerial reconnaissance

Israel has unveiled the Oron, not just a new aircraft, but a breakthrough aerial reconnaissance and control concept. Created in the face of military pressure and technological race, this multi-purpose aircraft combines advanced technology, modular architecture and artificial intelligence, forming a new standard for future intelligence platforms.

Oron is the result of decades of development, starting with legendary systems such as Shavit and Eytam.

From "Shavit" and "Eytama" to a single multifunctional platform

Previously, Israeli intelligence relied on specialized platforms: Shavit provided strategic interception of signals, and Eitam acted as a "flying command center" linking military units. However, these functions were performed separately, which slowed down the processing of information.

Oron is changing the paradigm: it combines electronic intelligence (RER), visual surveillance, analysis and control into a single system. Now a single platform can do everything from data collection to strike coordination.

Situational Fabric: How Oron turns Data into a Strategy

The key feature of Oron is the formation of the so—called "situational fabric". There are several teams of analysts working on board: some are engaged in signal interception, others in image processing, and others in radar and electronic mapping. All data is integrated into a single network and processed in real time.

Artificial intelligence systems automatically detect anomalies, identify sources of threats, and make predictions. For example, algorithms can detect signs of preparation for the deployment of troops or the launch of drones long before it becomes obvious to traditional intelligence.

Gulfstream 550 as a base: stealth, range, economy

The Gulfstream 550 business jet became the technical base of Oron. Its choice is due to a number of advantages: high flight range (up to 12,000 km), the ability to use short runways, low radar visibility and lower operating costs compared to heavy military aircraft.

This allows Oron to control vast territories — from the Middle East to North Africa, the Caucasus and parts of the Indian Ocean. The civilian platform also simplifies modernization: the modular electronic architecture allows for the rapid adoption of new technologies.

Role in combat and political operations

From a military point of view, Oron transforms the Israeli Air Force into a more "connected" and reactive force. The aircraft can coordinate the actions of fighters, air defense systems, and ground units, providing them with up-to-date intelligence that instantly transforms into combat instructions.

Politically, Oron strengthens Israel's position as a key partner of the United States and NATO. Its technological model — compact, efficient and scalable — can become a model for other countries. In addition, the possession of such a platform increases the value of Israel in intelligence alliances, especially in the context of confrontation with Iran and its proxies.

The Godfather of AI: Autonomous weapons simplify warfare by turning soldiers into "dead robots"

Jeffrey Hinton warns that deadly autonomous weapons and AI make it easier to start wars by reducing casualties among soldiers.

Jeffrey Hinton, the acknowledged godfather of artificial intelligence, noted that the development of lethal autonomous weapons greatly simplifies the conduct of military operations. According to him, the use of killer robots is radically changing military ethics and practices, turning soldiers into "dead robots" instead of the traditional "body bags."

AI has already significantly transformed the battlefield with the help of smart drones and other autonomous systems. Although the use of AI can reduce human losses, it is this factor that causes concern for Hinton, since facilitating the conduct of war increases the risk of its outbreak.

According to Hinton, the introduction of autonomous systems will be an unexpected opportunity for weapons manufacturers. States may experience fewer moral and political obstacles to the use of force, which may lead to an increase in the number of conflicts and a change in global security.

DARPA and five American companies are developing "strange" covert communication systems

PWND2 Program: when "strangeness" is an advantage

The US Department of Defense Advanced Research Projects Agency (DARPA) has launched a large-scale collaboration with leading American technology companies under the PWND2 program (Provable Weird and Novel Deployments 2 — "Deployment and Detection of Provably Strange Networks, Phase 2"). The goal is to develop fundamentally new communication systems, whose signals will be so "strange" and atypical that modern electronic intelligence systems simply will not be able to recognize or classify them.

This approach turns the unusual into a defense.

How provably strange networks work

Traditional methods of communication masking — encryption, frequency hopping — are increasingly amenable to analysis using AI and big data. PWND2 offers a different way: creating networks whose signatures (radio prints) deliberately differ from any known patterns, whether it's Wi—Fi, cellular networks, or military channels.

These "provably strange" signals are not just encrypted — they look like noise or an anomaly to an outside observer, which makes them virtually invisible to detection and analysis systems. Even with interception, it is impossible to decode them without knowing the unique "strategic key" of network behavior.

Key participants and contracts worth $ 29.2 million

In July and August 2025, DARPA signed contracts with five American companies.:

• Stealth Software Technologies (Arizona) and SRI International (California) are the original developers of the concept, specializing in cryptography and AI. The contracts are worth a total of $7.4 million.
• Two Six Labs (Virginia), RTX BBN Technologies and Systems & Technology Research (both in Massachusetts) joined later with contracts worth $21.8 million.

Each company develops its own approaches to the implementation of "strangeness" — from dynamic protocols to adaptive AI-controlled waveforms.

Countering modern surveillance systems

In an environment where the enemy has powerful radio monitoring and machine learning tools capable of detecting hidden signals, traditional encryption methods are no longer sufficient. PWND2 offers a radical solution — not to hide the signal, but to make it so alien that it ceases to be an object of interest.

If the program is successful, such networks can be used to connect intelligence groups, drones, and command centers in high-risk interception environments.

American unmanned fighter satellites: launch of flight tests of the CCA program

The YQ-44A and YFQ-42A prototypes are ready for their first flights

According to Pentagon sources, the launch of flight tests under the Collaborative Combat Aircraft (CCA) program is just around the corner. It is expected that at least one, and possibly both, prototypes will take to the air in the coming days — the YQ-44A Fury from Anduril Industries and the YFQ-42A from General Atomics. This will be the official start of the flight test and evaluation phase, which will lay the foundation for future mass production contracts.

A representative of the US Department of Defense said that both vehicles are in the "practical launch readiness" stage, which means the completion of ground checks, system tests and preparations for the first flights. Successful tests will allow the USAF to identify the most promising platform for further development.

The essence of the Collaborative Combat Aircraft (CCA) program

The CCA program aims to create so-called "unmanned fighter satellites" — drones capable of escorting fifth-generation manned fighters such as the F-35 and F-22, and in the future, the sixth generation (NGAD). These unmanned platforms will perform a wide range of tasks, from reconnaissance and electronic warfare to strikes against ground and air targets.

The key feature of CCA is the use of artificial intelligence for autonomous maneuvering and decision—making in combat, while control remains with the pilot operator in the lead aircraft. This makes it possible to significantly expand the combat zone, reduce risks for the crew and increase the overall efficiency of the aviation group.

From testing to mass production

The flight test phase will last for several years and will include testing for endurance, maneuverability, compatibility with manned platforms and resistance to interference. The final decision on the selection of the winner and the conclusion of contracts for mass production is planned to be made in fiscal year 2026.

The first serial production of the CCA is expected to enter service with the US Air Force in the late 2020s and early 2030s. This will be an important stage in the transformation of aviation from a predominantly manned to a hybrid system where man and machine work in close cooperation.

The role of artificial intelligence and the future of manned aviation

The CCA program marks the transition to a new paradigm of aerial combat. Artificial intelligence allows drones to operate in difficult conditions where human reaction may not be fast enough. At the same time, the pilot retains strategic control, delegating tactical tasks to his "unmanned partners."

This approach not only enhances combat capabilities, but also makes aviation more resilient to losses. CCA drones can be used as expendable platforms, allowing you to save valuable manned aircraft and crews.

China is preparing to display new combat drones at a military parade

Before the upcoming major military parade in Beijing, the first footage and data on China's new combat drones have appeared. At least five different models of unmanned aerial vehicles belonging to the loyal wingman class are visible on satellite images and in training footage. These vehicles are similar to what the US Air Force designates as joint combat aircraft (CCA — Collaborative Combat Aircraft).

The new Chinese drones are characterized by low visibility, autonomy and the ability to operate in close cooperation with manned fighter jets. They are designed to perform a wide range of tasks, from reconnaissance and electronic jamming to strikes against ground and air targets.

It is expected that some of the models presented at the parade may be advanced unmanned combat aircraft (UCAV) with increased autonomy, including the ability to make decisions in the air without constant operator control. At the same time, they retain the ability to accompany and support manned aircraft, enhancing their combat effectiveness.

The demonstration of such technologies highlights China's desire to reduce the technological gap between leading countries in the field of advanced aviation and artificial intelligence. The display of the new UCAVs at the parade is not just a spectacular event, but a signal of China's readiness to move to a new level of digital and autonomous aerial warfare.

Hackers and defenders in the age of AI: who will deceive whom?

Artificial intelligence (AI) has long ceased to be just the technology of the future — it is already here, and is actively used in one of the most intense digital confrontations: between hackers and cybersecurity experts. In fact, both have entered into a real arms race, where the main trump card is the ability to learn faster, disguise and attack.

Previously, cyber attacks were often based on patterns: phishing emails with grammatical errors, fake websites with suspicious addresses. Now everything is different. Attackers use AI to generate phishing messages that perfectly mimic the communication style of colleagues or management. These emails sound natural, don't arouse suspicion, and work much more effectively.

But this is just the beginning. AI helps hackers:

- Automate the search for vulnerabilities in systems.

- Create fake voices and videos (deepfakes) to deceive employees or gain access.

- Disguise malware as legitimate programs.

- Adapt real-time attacks to bypass defenses.

On the other hand, companies responsible for cybersecurity are also not sitting idly by. They use AI to:

- Analyzing huge amounts of data and detecting anomalies.

- Predicting attacks based on network behavior.

- Rapid response to incidents — sometimes faster than a human can do.

It turns out that both "criminals" and "good guys" use the same technology. There is a difference in goals: some want to steal data, paralyze systems, or extort money, while others want to protect infrastructure, business, and privacy.

Of particular concern is the use of AI by foreign intelligence agencies. Government hacker groups are increasingly using AI to conduct targeted, covert, and long—term operations ranging from industrial espionage to interference in critical systems.

As a result, the smarter the AI gets, the higher the stakes. Now the struggle is not only between people, but also between algorithms. And the one whose AI learns faster, adapts faster, and anticipates the opponent's moves will win.

Five military robots with artificial intelligence already used on the battlefield

 The five military robots with artificial intelligence already in use on the battlefield include:

1. The MQ-9 Reaper is an American attack drone used by the US Air Force and Navy, as well as the armies of Great Britain, Italy, France and other countries. It is capable of carrying out reconnaissance missions and striking with AGM-114 Hellfire missiles and bombs with laser or GPS guidance. The MQ-9 Reaper works autonomously and allows you to make lightning-fast decisions without delay from the operator, significantly reducing the risk to soldiers and increasing the effectiveness of combat missions. Its first flight took place in 2001, it has been in operation since 2007, and flight autonomy reaches 14-27 hours, depending on the modification. As of 2018, there were approximately 222 such devices in the US Air Force.

2. Uran-9 is a Russian combat robot with artificial intelligence designed for combat operations in urban and suburban areas. It is equipped with various weapons and surveillance systems.

3. THeMIS is a robotic platform used for combat, reconnaissance and support missions. The main application is infantry support and dangerous missions.

4. Indian Robotic Mule is a robotic truck or porter that helps transport goods and performs auxiliary functions on the battlefield.

5. Goalkeeper CIWS is an automatic air defense system that includes artificial intelligence elements to identify and defeat aerial targets with high reaction speed.

These AI-powered robots make it possible to process sensor data, avoid threats, and operate as efficiently as possible in combat, which marks a new era in warfare. Every year, more and more countries are expanding the use of such systems for various combat, border and auxiliary tasks, reducing risks to people and improving the operational capabilities of the army. AI in military robots allows you to make split-second decisions when time is critical.

Leonardo DRS has successfully conducted the first tests of a marine package of anti-drone equipment on small surface vessels

On August 12, 2025, Leonardo DRS announced the successful completion of the first series of open water tests of its advanced marine equipment package, the Counter-Unmanned Aerial System Maritime Mission Equipment Package. This system is specifically designed for installation on small unmanned surface vessels (USVs) and is designed to protect naval fleets from threats from unmanned aerial vehicles.

The new system is a scalable solution based on Leonardo DRS's expertise in ground-based mobile drone and short-range air defense systems. It integrates active and passive radio frequency (RF), optoelectronic, infrared (EO/IR) sensors, 4G/5G-based electronic countermeasures systems, and a variety of kinetic and non-kinetic weapons.

The tests were conducted in realistic marine conditions and confirmed the system's capabilities in detecting, identifying and tracking aerial threats, as well as tracking surface objects. Special attention was paid to working in a complex electronic environment with the BlackLab passive radio frequency detector and the STAG EO/IR platform with thermal imaging cameras.

The system is controlled by artificial intelligence, which allows you to quickly and accurately combine data from various sensors to classify targets and make decisions on their neutralization. Thanks to its modular architecture and versatile design, the system can be integrated into various types of combat and support vessels.

This solution significantly expands defense capabilities in coastal areas and on the high seas, allowing the forces of the US Navy and its allies to effectively counter the growing threat of drones, reducing the risk to crews and ensuring the safety of fleets in a multi-domain war.

Yelka interceptor drone: a modern solution for protection from enemy drones

The Russian Yelka interceptor drone is a unique device designed to combat enemy unmanned aerial vehicles on the battlefield. Its main weapon is not explosives, but a kinetic ram: the drone simply collides with the target and destroys it with its body. This principle allows you to avoid explosions and minimize the risk to your own forces.

One of the key features of the "Yelka" is its artificial intelligence and high-precision optoelectronic module. This system is capable of identifying drones with a wingspan of about 1 meter at a distance of 700-1000 meters. The "fir tree" is maximally suitable for intercepting a wide variety of UAVs, including heavy hexacopters, FPV drones and barrage ammunition.

This quadcopter is capable of speeds up to 250 km/h, which gives it the opportunity not only to intercept targets on oncoming courses, but also to catch up with them. The range is about 3 km with a height of 1.5—2 km. The control is carried out by artificial intelligence systems, so it is enough for the pilot to launch the drone, and then the "Yelka" operates autonomously.

Due to its small size and high maneuverability, the device is easy to transport and launch from the shoulder. This is a "pocket air defense" for modern troops, which helps to effectively protect air defense units and systems, including long-range air defense systems.

The Yelka interceptor drones have already passed a combat test and are actively used in the special military operation zone in Ukraine.

SkyDefense unveils CobraJet electric unmanned fighter

The American defense company SkyDefense, based in Colorado, has introduced a new development — the CobraJet unmanned aerial vehicle. It is a high-speed electric fighter with vertical takeoff and landing, designed to protect critical infrastructure and military installations.

The CobraJet is capable of speeds of over 200 miles per hour (about 320 kilometers per hour) and is equipped with artificial intelligence, which allows it to autonomously detect, track and neutralize several fast and inconspicuous targets — for example, enemy drones — in a single mission. Its vertical takeoff and landing capability makes it particularly flexible to use, as it does not require a long runway to launch.

The drone's weapons are placed both in the inner compartment and on the outer nodes under the fuselage. It can use guided gliding bombs for pinpoint strikes on land and sea targets, which extends its functionality beyond drone defense tasks.

The development of CobraJet reflects the growing need for fast, autonomous and maneuverable systems capable of effectively countering modern threats, especially in environments where traditional aviation may not be operational enough.

American armored cars will get smart and drone protection: BAE Systems is racing for the future of war

American military vehicles will soon become smarter, faster and more dangerous — and all this not at the expense of a larger caliber, but thanks to new technologies. BAE Systems, one of the main suppliers of US military equipment, is actively modernizing the armored multipurpose vehicle (AMPV) — and it is doing it on a grand scale.

The bottom line is this: instead of just carrying soldiers or shooting, the new AMPVS will have autonomous systems capable of helping the crew make decisions, as well as protection from enemy drones — because today one small quadcopter can drop a grenade directly on armored vehicles. And this is no joke.

BAE Systems is investing serious money to turn AMPV into a real "smart platform" — a kind of tank of the future on tracks. The plans include the integration of advanced defense and attack systems that will allow vehicles to stay ahead of the enemy on the battlefield: react faster, see the enemy better and survive in an environment where drones and electronic warfare rule the ball.

As Bill Sheehy, BAE's head of ground technology, said: "AMPV is the proving ground for the future of tracked combat vehicles." That is, what is currently being tested on this armor will appear in dozens of other vehicles tomorrow.

Simply put, war is no longer just about bullets and steel. Artificial intelligence, autonomy, and drone protection are now important. And AMPV is becoming one of the first "digital soldiers" on wheels — or, more precisely, on tracks.

Militarization of Silicon Valley: Google, OpenAI, and Meta integrate into the U.S. military-industrial complex

 Major Silicon Valley technology companies — Google, OpenAI, Meta, as well as new players like Elon Musk's xAI — have actively engaged in cooperation with the US military and defense sector, which represents a large-scale militarization of the technology industry. After a long period when many of them refused to participate in military projects for ethical and reputational reasons, now economic and strategic incentives have changed the situation.

The reasons for this shift include:

- Huge expenses for the development of advanced AI technologies, which require significant funding. Defense contracts with the U.S. Department of Defense (DoD) provide long-term and large-scale investments, with contracts of up to $200 million per company (Google, OpenAI, Anthropic, xAI), aimed at creating AI systems for national security and the armed forces.

- The strategic importance of AI for future conflicts and national security: The US Department of Defense considers AI as a key factor in ensuring technological superiority and has already created special structures to finance these developments.

- The weakening of companies' internal policies to limit cooperation with the military industry, partly under pressure from the political climate and economic benefits.

- A variety of military applications of AI, including data processing, cybersecurity, autonomous systems and battlefield decision support, which expands the capabilities of technology and gives companies new markets.

- Joint integration initiatives and pilot projects with military departments aimed at accelerating the implementation of AI in real military tasks.

Thus, the "militarization of Silicon Valley" is a shift in the relationship of technology giants with the military sector, where previously principled detachment was replaced by active involvement in defense—related projects. This reflects economic pressures, the strategic agenda, and the changing landscape of the AI and technology industry.

Microphones against hypersonic missiles: how an old trick from the First World War can help against new weapons

Hypersonic missiles are a radar nightmare. They fly 5-10 times faster than sound, are enveloped in plasma, which "extinguishes" radio signals, and change their trajectory on the way to the target. But there is one caveat: they make a lot of noise. And this is exactly what a new, but essentially ancient idea is based on — to detect missiles not with radars, but with microphones.

Does that sound like a joke? But it's not. It's a return to the roots. Back in World War I, armies used arrays of microphones to determine where enemy artillery was firing from. The time difference between the sound of the shot coming to different microphones was measured, and the launch point was calculated from this difference. Today's technology just took this idea and put smart glasses on it.

Modern acoustic systems are not just microphones, but entire networks of "smart" sensors with integrated processing. They determine the direction of the sound themselves, classify it (it's a drone, it's a jet engine, it's an explosion), and immediately send the data to the command center. And they do it quickly, cheaply and without radiation, which means they are difficult to detect.

Infrasound sensors are especially powerful — they catch waves with a frequency below 20 Hz, which spread over hundreds of kilometers, almost without losing strength. Such systems are already in use: for example, the American UTAMS tracks missile launches, mines and shelling by sound. And in Ukraine, the Zvook network operates — thousands of acoustic sensors that use artificial intelligence to analyze noise and help air defense to find drones, Grads and helicopters. It's not a replacement for radars, but it's a great addition, especially when the enemy uses silencers and low-profile targets.

So can such a "microphone shield" stop a hypersonic missile?

Partly— yes, but not in the way it seems.  

While the rocket is at launch, during takeoff, it emits a powerful engine roar and an explosion — this can be detected. Acoustic networks can record the moment of launch and orient other systems: satellites, IR sensors, radars. That is, the microphones here are not the main character, but the "starting whistle".

But as soon as the rocket goes into the stratosphere, the sound lags behind. A sound wave travels at 340 m/s, while hypersound travels at 1,700 m/s or higher. In addition, the air is thin at high altitude, and sound does not spread well. Therefore, acoustics are powerless in the terminal phase of flight.

The real "hunters" of hypersound are: 

- Satellites with infrared sensors that see the fire plume from the engine;  

- Powerful AFAR radars capable of operating in interference conditions;  

- Over-the-horizon systems that track missiles beyond radio visibility;  

- And most importantly, data fusion systems that combine information from all sensors into a single picture.

So microphones are not a panacea, but a valuable element of the mosaic. They are cheap, easy to deploy, and work great in the near-term. In the fight against hypersound, they will not replace space systems, but they can provide those very first seconds that decide the outcome of an attack. And in war, as you know, every second counts.