General Algorithm

Artificial intelligence has confidently become an indispensable assistant in many areas that improve people's quality of life — from automating internet search to education and medicine. However, it has, like any technology, revealed another side.

Recently, engineers from leading corporations are resigning in protest against the use of their developments in the military-industrial complex, musicians are removing entire discographies from streaming platforms that finance military AI, and news that an autonomous drone independently made and executed the decision to destroy an enemy is no longer surprising anyone.

ForkLog examined how, in just 10 years, artificial intelligence transformed from a futuristic project into a part of everyday reality for armies around the world, who leads in the new arms race and what the militarization of neural networks may lead to.

How Artificial Intelligence Became a Weapon

In the early 2000s, the American agency DARPA began experiments with autonomous systems that allowed machines not just to execute commands, but to make decisions independently on the battlefield. Prototypes like Crusher — a six-ton reconnaissance robot — could navigate rough terrain without human involvement. At the same time, experiments were conducted with the automatic sniper drone ARSS. These developments became the first building blocks in the foundation of military AI.

In April 2017, the U.S. Department of Defense launched Project Maven — a large-scale program for the combat use of artificial intelligence. Its task was to analyze video streams from drones using machine learning algorithms. The program rapidly developed, and just a few months later, the technology was added to Predator and Reaper drones.

Inside the industry, this caused shock: thousands of Google employees — the project's main partner — signed a petition against the company's participation in military initiatives, and dozens of engineers resigned. In 2018, the corporation officially refused to continue cooperation.

In the Second Nagorno-Karabakh War, Azerbaijan widely used Turkish drones Bayraktar TB2 and Israeli Harop, as well as experimental kamikaze drones Kargu-2. These systems incorporated elements of autonomy, such as targeting, recognition of equipment, and tracking of targets. However, the exact level of independent decision-making, including the ability to attack without an operator, is not officially disclosed and remains a subject of expert assessments and assumptions. Nevertheless, this conflict demonstrated the extensive use of modern drone technologies based on AI.

China, for its part, is actively developing a strategy for military-civil fusion. This policy combines the efforts of the defense sector, academic institutions, and technological giants. Chinese investments are focused on the development of autonomous systems for monitoring, interception, reconnaissance, and maritime operations. The main emphasis is on making artificial intelligence not only a weapon but also a key infrastructure of the modern military.

Meanwhile, in Israel, AI developed along a parallel course. Since 2021, the military began actively implementing artificial intelligence systems for automating targeting in the Gaza Strip. These algorithms analyzed data arrays — phone calls, movements, digital activity — and helped form lists of targets for strikes. As a result, the Lavender and Gospel systems emerged, which were used in large-scale operations in 2023. This sparked international criticism: human rights activists and the UN questioned the legality of strikes taken based on AI recommendations.

The real war of algorithms began in 2022, with the onset of the full-scale conflict between Russia and Ukraine. From that moment on, AI became a true participant in the hostilities. The American company Palantir provided Ukraine with tactical platforms for intelligence analysis and strike planning.

At the same time, Ukrainian startups, including Gogol Brain, were developing AI modules for drones, including semi-autonomous FPVs capable of identifying targets and operating in electronic warfare conditions. Artificial intelligence has become not an auxiliary tool, but the core of the reconnaissance-strike system.

Russia is using Iranian Shahed-136 drones, modified for increased accuracy and resistance to electronic countermeasures. Some sources report the possible integration of AI into these drones, but there is no official confirmation of this.

In 2025, the world is no longer discussing whether AI will become a weapon — it has already become one. Taiwan has launched a national program to produce 25,000 FPV drones with AI navigation, creating a layered network of autonomous defense against a potential invasion from China.

India, also considering the growing tension along the border with China and Pakistan, has begun to implement artificial intelligence in security systems. Since 2022, about 140–145 AI-supported surveillance platforms have been deployed, integrating cameras, thermal imagers, radars, and drones for facial recognition.

In just 10 years, artificial intelligence has evolved from auxiliary software for video analysis to autonomous weapons conducting reconnaissance, selecting targets, and often pulling the trigger. The world has entered a new era, and this process cannot be stopped.

How much is being invested in the "smart" war?

It is almost impossible to know exactly how much and on what the military spends, especially in the field of AI, as most programs are classified. However, it is possible to establish at least the general outlines based on open data.

The Volume of the Global Defense AI Market

Data: author. The global military AI market was estimated at $4.8 billion in 2022 and $9.3 billion in 2024. According to the most conservative scenario, the defense artificial intelligence market will grow to $13–19 billion by 2030, while a moderate scenario suggests growth to $25–30 billion, and with further escalation of military conflicts around the world, growth may reach $35–40 billion and above. At the same time, up to 50% of investments may come through the private sector: startups, venture funds, and big tech.

Data: author USA

In the race for defense AI systems, the USA currently leads. This is largely due to the close cooperation between the Pentagon and major technology companies: Palantir, Anduril, Shield AI, Scale AI are receiving contracts worth billions of dollars. In 2023, the Ministry of Defense created Task Force Lima — a division for the development and implementation of generative AI in the army, as well as a large-scale program called Replicator.

The Pentagon's budget allocated for AI technologies has increased from $600 in 2018 to $1.8 billion in 2024, while the total costs for military artificial intelligence are estimated at $4–6 billion annually.

China

Through the policy of "military-civil fusion," the largest AI companies—SenseTime, Megvii, iFlytek—receive government support for research and development work. According to CIGI, in 2023 the funding volume for dual-use technologies in China amounted to $2.5–3 billion per year.

China is actively developing autonomous drones of the Wing Loong series (, CH-5), and underwater unmanned systems. Tests of AI fleets, including the use of drone swarms, are being conducted in the South China Sea. Additionally, artificial intelligence is being actively integrated into cyber operations, air defense systems, ISR, and forecasting.

Russia

The Russian Federation is betting on mass production and actively developing cheap UAVs, AI for munitions, electronic warfare systems, and autonomous navigation. Most of the expenses are classified, but independent estimates indicate $300–500 million per year (2023). The growing share of AI integration is confirmed by an increase in Russia's overall military budget (+30% in 2025).

EU and the United Kingdom

Since 2021, the EU has invested €1.5 billion through the European Defence Fund in projects related to AI, robotics, and autonomy. According to the 2025 Defence Review, artificial intelligence is recognized as one of the most important components of the future army. Key areas include target recognition, navigation without GPS, and compatibility of AI systems among NATO member states. Autonomous systems, threat analysis modules, predictive analytics, and countermeasures against enemy drones are being tested. The UK is developing defence AI hubs and closely cooperating with the USA and Israel. Full integration of AI into the armed forces is planned by 2030.

Israel

One of the few countries already using AI in real combat operations

time, including the integration of artificial intelligence into missile defense systems, drones, ground complexes. Estimated budget: $0.5–1.5 billion per year.

India

India is betting on partnerships with private companies. In 2024, investments in defense AI amounting to $750 million have been announced until 2027. An Artificial Intelligence and Robotics Center has been established, and drone tests are being conducted. In addition, AI platforms for unmanned logistics and reconnaissance are being tested in the Himalayas.

Ukraine

In 2022, Ukraine became the first country to widely use AI and autonomous technologies in the context of full-scale war. In 2024, the government announced the creation of a defense AI technology center in collaboration with international partners. However, the state budget for these tasks remains limited, with various estimates ranging from $200–400 million per year.

There is also extensive collaboration with startups and volunteer projects in AI analytics, computer vision, and tactical navigation. Despite limited funding, flexibility, speed of implementation, and field adaptation make the Ukrainian experience unique — it is studied worldwide, including by NATO.

South Korea

South Korea is developing predictive threat analytics and military intelligence systems. AI turrets SGR-A1 have already been installed on the border with North Korea, operating in semi-automatic mode ( with human target confirmation ). Annual spending on defense artificial intelligence is estimated at $500–600 million with plans to increase to $700 million by 2025.

Taiwan

Taiwan is preparing for possible defense in case of a Chinese invasion, building infrastructure for mass production of AI-controlled drones. Investments in

defense artificial intelligence accounts for between $600 million and $1 billion per year

Main Directions of Defense AI

Unmanned Aerial Vehicles. AI is used for autonomous navigation, target acquisition, threat identification, and swarm organization of drones. It is utilized in reconnaissance, strikes, electronic warfare, and surveillance. Drones coordinate with each other like bees. One unit can be easily shot down, but a swarm is almost invulnerable.

Autonomous control systems for combat platforms (land, sea, air). AI manages tanks, combat UAVs, unmanned ships.

Military logistics and supply. Models predict ammunition shortages, optimize supply routes, manage equipment repairs in the field. Route optimization, predictive maintenance of equipment, warehouse automation. Used to accelerate response and reduce costs.

Intelligence, surveillance, and data analysis. AI processes satellite, radar, video, and audio data, including information from open sources, and also applies facial and behavioral recognition systems, behavior analysis, and threat forecasting. Now satellite data is analyzed by machine vision and predictive analytics systems. Previously, analysts spent hours analyzing images; today, it takes seconds.

Cybersecurity and Cyberwarfare. Adaptive protection of air defense systems, UAVs, and communication systems. AI is used both for detecting and blocking cyberattacks and for carrying them out. Artificial intelligence can hack, protect, and substitute data faster than a human. Cyberattacks no longer require an army of hackers — they are carried out by models.

Command and Control. AI helps analyze the operational situation, recommend actions, and model scenarios. It is used to support command decisions, not to replace humans.

Combat algorithms and guidance systems. Target identification and tracking, ballistic calculation, assistance in target designation.

Simulations and military exercises. Training personnel in a simulated environment, generating combat scenarios, predicting enemy behavior. It is also used for testing new strategies.

Psychological operations and information warfare. AI is used for generating deepfakes, analyzing social media, and targeted misinformation. It is used to influence the opinions of both the civilian population and military personnel.

AI for soldier support. This includes "smart" systems to assist infantry: AI helmets, exoskeletons, HUD interfaces, individual combat companions.

Satellite Management. Using AI to track objects in space and for autonomous piloting. Example: the American SDA system with elements of artificial intelligence.

Early warning and automatic response nuclear systems. Strategic systems with AI analysis elements and strict autonomy limitations. Used only as an assistant, full autonomy is not allowed due to escalation risks.

Neural Networks for Diplomacy. Predicting the reactions of countries to various strikes, threats, or cyberattacks is a new type of AI that operates at the intersection of military logic and psychology.

Instead of conclusion

A new world war is already underway, simply in the form of algorithms fighting for control over land, water, air, space, cyberspace, and the consciousness of people. The main difference is that there is now less infantry, but more technological solutions. And the further we go, the more these solutions will be made not by generals, but by machines.

AI is no longer the future of war, but its present. And although China and the USA are the leaders of the race, more and more countries are joining the game. In 2023 and 2024, the REAIM summits took place in The Hague and Seoul. More than 50 countries signed a declaration on the responsible use of military AI, and the UN is calling for a ban on fully autonomous systems that can kill without human involvement.

The main question is whether the international community will manage to introduce regulations before AI writes the rules itself.

Text: VGI666