One nation is currently conducting the world's most sophisticated warfare: Israel. The shocking surprise attack during the 1973 Yom Kippur War fundamentally transformed the country. Driven by the desperate resolve of 'Never Again,' their journey has culminated in an AI military revolution that is now the focus of global attention.

The Israeli military is evolving at a staggering pace. In the past, identifying 50 targets per year was considered a limit; today, AI systems seamlessly identify over 100 targets in a single day. Much like a supercomputer playing a game of chess or Go, artificial intelligence oversees the entire battlefield, calculating the optimal moves.

Elite units such as 'Talpiot (Talpiot)' and 'Unit 8200' are at the heart of Israel's technological prowess. By recruiting the nation's brightest youth and training them in cutting-edge technologies, Israel has created a perfect virtuous cycle where these individuals later found startups that supply advanced technology back to the military. The establishment of the 'AI and Autonomy Administration' in 2025 demonstrates a powerful national commitment to systematizing these efforts.

AI systems like 'Habsora (Gospel)' and 'Lavender' have completely rewritten the rules of engagement. By analyzing vast amounts of data, they can generate thousands of strike targets in an instant, achieving operational speeds previously unimaginable for humans alone. While 'Lavender' identifies up to 37,000 individuals as targets with a claimed 90% accuracy, its 10% error rate and the 'Where's Daddy?' system—which targets individuals specifically when they are at home with their families—have sparked intense international controversy.

With the deployment of these systems in the Gaza Strip, humanity has witnessed the 'first true AI war.' Starting with Operation 'Guardian of the Walls' in 2021 and continuing through Operation 'Iron Swords' in 2023–2024, AI played a central role in attacking over 12,000 locations in just 27 days. However, these advancements have faced sharp criticism for threatening the fundamental principles of international law, citing the limits of AI accuracy, the 'rubber-stamp' nature of human control (where targets are approved in as little as 20 seconds), and pre-determined high thresholds for civilian casualties.

1 Strategic Background of Israeli Military AI

1-1 National AI Strategy and Military Priorities

Israel's military AI is not merely about catching up with technology; it was a desperate choice for survival. The surprise attack of the 1973 Yom Kippur War left a painful lesson, marking the moment when the obsession with securing a 'qualitative military edge through technological superiority' began.

As irregular armed groups like Hamas and Hezbollah emerged and the nature of warfare became more complex, Israel sought new solutions. They bet everything on information processing capabilities, data analysis, and algorithmic superiority rather than weaponry alone. AI became the core technology for analyzing massive amounts of surveillance and reconnaissance data in real-time, rapidly locating hidden enemies, and theoretically minimizing civilian casualties in a transformed battlefield.

1-2 The Israeli Defense Industry and AI Ecosystem

Israel's military AI ecosystem is among the most unique and powerful in the world. It operates through a perfect synchronization of state-led talent cultivation and private sector innovation.

The 'Talpiot' program selects a small number of scientific prodigies each year for nine years of intensive training. These individuals participate in the development of core weaponry like the Iron Dome and, after their service, lead the Israeli tech ecosystem through the 'Talpinet' network. The bond is so strong it is often said, 'Once Talpiot, always Talpiot.'

'Unit 8200' is Israel's premier intelligence unit, often compared to the U.S. NSA. It produces top talent in signals intelligence, cyber warfare, big data analysis, and machine learning. After discharge, Unit 8200 alumni create successful startups like Wiz and NSO Group, contributing back to national defense in a virtuous cycle.

Specialized units such as Unit 81 (the technology unit), LOTEM, Unit 3060 (operational and visual intelligence systems), and the Matzpin unit (developers of the Genie chatbot) handle military AI research and development in their respective domains.

The defense industry itself is remarkably diverse. State-owned enterprises like IAI (which operates an AI Center of Excellence) and Rafael (developers of Iron Dome and Spike), along with the publicly traded Elbit Systems, account for 90% of defense exports and lead AI technology development. As of 2025, the number of Israeli defense companies has surged to 312.

Innovation from startups is also remarkable. Companies like Corsight (facial recognition), Kela (battlefield AI platforms), Deep Instinct (cybersecurity), AI21 Labs (natural language processing), and Run:ai (GPU optimization, acquired by NVIDIA) develop technologies applicable to both civilian and military sectors.

Global corporations also collaborate closely with Israel. Palantir provides data analysis platforms, while Microsoft, Google (Project Nimbus), Amazon, and OpenAI support cloud computing and AI models. The fact that these companies supported Israeli military operations even during the Gaza war demonstrates the blurring lines between technology and modern warfare.

2 Core AI Systems

Israeli military AI spans the entire process of warfare, from intelligence gathering to actual strikes. Their 'algorithmic arsenal' has innovatively transformed the speed and scale of the battlefield.

2-1 Target Identification and Strike Systems

Habsora (Gospel): Named with religious overtones, this system literally brings 'good news' in a military context. It analyzes all types of data—including drones, satellite imagery, and signals intelligence—to locate military facilities such as armories, command posts, and tunnels. Consequently, the Israeli military increased its target identification from 50 per year to over 100 per day, earning the system the nickname 'the target factory.'

Lavender: A signature AI system designed for targeting individuals. Using machine learning, it analyzes visual data, mobile phone information, social media, and call records to score the probability (from 1 to 100) of an individual being a member of Hamas or PIJ. It has generated a list of up to 37,000 potential enemies in Gaza. While boasting 90% accuracy internally, its 10% error rate carries the risk of innocent civilians becoming targets.

'Where's Daddy?': This chillingly named system tracks when a targeted individual identified by Lavender enters their family home. It has faced severe criticism for its ruthless strategy of striking when the target is with their family.

Fire Factory: An operational optimization system that analyzes past mission data and current target information to calculate necessary munitions, suggest optimal strike times, and assign priorities and strike platforms for thousands of targets.

2-2 Intelligence Collection and Analysis

AI Utilization in Unit 8200: Enhances signals intelligence and cyber warfare capabilities through large-scale data analysis, machine learning, and natural language processing. It is used to thwart terrorist attacks, track enemy locations, and analyze social media and communication data.

Video Analysis and Surveillance Systems: AI installed on border surveillance cameras automatically identifies and alerts to the presence of people, vehicles, or individuals carrying weapons through real-time video analysis.

Genie: A military chatbot developed by the Matzpin unit of the Communications Corps. It assists commanders in decision-making by analyzing real-time military intelligence and massive military databases. It is capable of analyzing multiple languages, including Hebrew, Arabic dialects, and English.

Alchemist System: Deployed as early as 2021, this system integrates diverse data into a single platform to identify threats like suspicious movements and rapidly alert combatants.

2-3 Defense and Interception Systems

Israel's multi-layered missile defense system is a premier example of the successful application of AI.

Iron Dome: A legend in short-range rocket interception. AI algorithms analyze real-time sensor data to detect, identify, and prioritize threats, supporting interception decisions. With a high success rate of over 90%, it also significantly reduces operational costs. It is considered a masterpiece of Rafael Advanced Defense Systems.

David's Sling: A medium-range missile and rocket defense system that, alongside Iron Dome, constitutes a multi-layered defense network. It utilizes AI for automated threat analysis and interception decision-making.

Arrow System: A long-range ballistic missile defense system where AI assists in trajectory prediction and calculating the optimal point for interception.

Edge 360: Developed by Axon Vision, this AI-powered system is installed on armored vehicles to detect potential threats from all 360 degrees and provide immediate warnings to the driver.

2-4 Command, Control, and Operational Support Systems

Integrated Command and Control Systems: By leveraging AI, these systems integrate various sensors and strike assets to display real-time battlefield intelligence on a single screen, enhancing situational awareness and accelerating commanders' decision-making speeds.

2-5 Autonomous and Semi-Autonomous Platforms

Drawing on extensive combat experience, Israel has developed a range of AI-driven military technologies, positioning itself as a global leader in the field.

Next-Generation Ground Combat Platforms

The Carmel Program is Israel’s futuristic armored fighting vehicle project, where AI plays a pivotal role. The AI can manage and execute a significant portion of driving and combat operations independently, without human intervention. It has advanced to a stage where it can navigate safely to a destination without a driver and respond automatically upon identifying an enemy.

The Barak tank is the Israeli Ministry of Defense's ambitious next-generation main battle tank. Its most innovative feature is the AI-driven 360-degree observation system. AI analyzes information flowing in from all directions in real time, providing the crew with comprehensive situational data. As if the tank were equipped with dozens of eyes, it monitors its surroundings without blind spots, and the fusion of data from various sensors by the AI has greatly improved target identification efficiency.

Innovation in Personal Weaponry

Smart Shooter is an AI-based optical sight mounted on rifles or machine guns used by individual soldiers. Functioning like the eyes of an elite sniper, it automatically detects and continuously tracks targets. When the shooter prepares to pull the trigger, the AI calculates and indicates the optimal firing moment, significantly enhancing accuracy. Recently, it has even been successfully deployed to down small drones, proving its effectiveness.

Autonomous Defense Systems

Semi-autonomous armed robots called Jaguar are deployed along the Israel-Gaza border. They provide continuous 24-hour surveillance, helping human soldiers avoid direct exposure in high-risk zones. While they operate under human command, they possess the intelligence to make basic independent judgments.

Guardium, developed by IAI, is a fully autonomous patrol system. It moves independently along pre-set routes, detecting and reporting any anomalies. Functioning as a robotic security guard, it maintains security by patrolling designated areas day and night.

AMSTAF, by IMI Systems, is an autonomous unmanned ground vehicle (UGV) capable of navigating to a destination without a human operator. It contributes to reducing human casualties by handling logistics transport and reconnaissance in dangerous areas.

Aerial Autonomous Weapon Systems

SkyStriker, developed by Elbit Systems, is an autonomous loitering munition (suicide drone) capable of independently locating and striking targets. Once launched, the AI detects and identifies targets within a specific area for precision engagement. Unlike conventional missiles, it can loiter in the target area and wait for the ideal attack opportunity, making it highly efficient.

3 Operational Deployment and Performance Analysis

3-1 AI Utilization in Gaza Strip Operations

The Gaza Strip conflict will be recorded as the first 'algorithm war' in history, where military AI systems were deployed by Israel on a massive scale in actual combat.

2021 Operation "Guardian of the Walls": Military leaders declared this operation the "first artificial intelligence war." They revealed that AI enabled the identification of 100 targets per day, a stark increase from the previous rate of roughly 50 targets per year.

2023-2024 Operation "Iron Swords": Following the Hamas surprise attack on October 7, the IDF struck over 12,000 targets in just 27 days using the 'Gospel' (Habsora) system. This illustrates the unprecedented scale and speed of AI-driven target generation systems. According to military press releases, the 'Gospel program' aided in the bombardment of 12,000 targets in Gaza. AI-based audio tools were also utilized to track the locations of high-ranking Hamas commanders.

3-2 Operational Effectiveness and Controversy

The IDF maintains that the implementation of AI systems enhances the precision and efficiency of the intelligence process, thereby minimizing civilian casualties. However, the reality on the ground has raised serious doubts about these claims.

Claims of Efficiency vs. Civilian Toll: While AI systems dramatically increased the speed and scale of operations, this has resulted in staggering civilian casualties. Since October 7, 2023, more than 50,000 people have died in Gaza and Lebanon, and nearly 70% of the buildings in Gaza have been destroyed.

The Paradox of Precision: Although AI is touted for its ability to provide more accurate target identification, testimonies suggest that rules of engagement were applied that permitted 15–20 civilian deaths for the elimination of a single low-level militant, and over 100 civilian deaths in the case of a senior commander. This implies an operational logic that prioritizes the 'efficiency of destruction' over precision, accepting massive collateral damage.

Accuracy Limits and Misidentification: The 10% error rate of the 'Lavender' system means that innocent civilians risk becoming targets. One intelligence officer testified, "I knew it was statistically accurate even if I wasn't certain the machine was right. So, I proceeded," highlighting the potential for misidentification due to errors in communication patterns or other factors.

3-3 International Legal Disputes and Ethical Issues

Israel’s use of military AI stands at the forefront of international legal and ethical debates, sparking serious concern within the international community.

The Autonomous Lethal Weapons Debate: On the international stage, including at the UN, Israel maintains that existing International Humanitarian Law (IHL) is sufficient to regulate new weapons systems, including autonomous lethal weapons, and that a new legally binding treaty is unnecessary. It argues that 'meaningful human control' is not an independent legal requirement of international law. However, Israel faces criticism for making de facto 'semi-autonomous killing' decisions by developing weapons with significant autonomy, such as the 'Harpy' suicide drone, and by operating AI systems in Gaza where human intervention is reduced to a mere formality.

Potential Violations of International Humanitarian Law:

The Principle of Distinction: The 'Lavender' system, with its 10% error rate, and the practice of targeting individuals at their homes threaten the principle of distinguishing between civilians and combatants.

The Principle of Proportionality: Pre-authorizing the deaths of 15 to 20 civilians to eliminate a single low-level combatant is criticized for neutralizing the judgment of proportionality between expected military advantage and collateral civilian damage, replacing it with a mechanical formula.

The Principle of Precaution: Approving AI-recommended targets in just 20 seconds and using unguided bombs makes it difficult to argue that all necessary precautions were taken to minimize civilian casualties.

Absence of Meaningful Human Control: While the Israeli military claims that AI systems are merely 'decision-support tools' subject to human approval, internal testimonies reveal that human analysts spend only '20 seconds per target' on review. Some analysts lamented that they added no human value beyond acting as a 'rubber stamp.' This highlights the risk of humans uncritically accepting AI outputs and losing effective control.

Accountability Gap: When AI algorithms function as 'black boxes' and human intervention is perfunctory, an 'accountability gap' arises, making it ambiguous who should be held responsible for illegal attacks. This could hinder the prosecution of war crimes and weaken the enforcement of international law.

Dehumanization of Warfare: Concerns have been raised that algorithm-based automated target generation and strikes increase the emotional and moral distance between commanders and violent outcomes. As targets are perceived as mere probability values in data, this process accelerates the dehumanization of war.

International organizations have expressed deep concern. The UN stated it is 'deeply concerned' by Israel's use of AI. The International Committee of the Red Cross (ICRC) emphasizes that 'meaningful human control' is essential in decisions to use force and that AI's unpredictability necessitates new norms and clear limitations. Human Rights Watch strongly criticizes the excessive reliance on AI tools, the lack of human oversight, and the use of flawed data and biased algorithms, arguing that Israel’s AI operations could constitute war crimes.

3-4 Future Outlook and Implications

Israel is pursuing a long-term national strategy to secure military AI superiority, but it simultaneously faces significant challenges.

The Quest for Sovereign AI: A core objective is establishing 'Sovereign AI'—the capability to independently develop and operate large-scale AI without relying on foreign commercial firms. In 2025, as part of the National AI Program, the construction of a large-scale AI supercomputer was announced. This demonstrates a strong commitment to providing computing resources across the AI ecosystem and avoiding technological dependence in future AI fields. The Israeli military's five-year AI plan (2024–2029) aims to build fully autonomous battlefield decision-making systems.

An Endless Arms Race: Hostile powers, including Iran, and neighboring countries are also pursuing the militarization of AI, fueling a continuous technological race. This exerts constant pressure on Israel to innovate while increasing the risk that advanced AI technology could spread to hostile non-state actors.

The Paradox of Over-reliance: The October 7, 2023, surprise attack by Hamas, which neutralized Israel's seemingly 'impenetrable' technological systems, taught a lesson that over-reliance on technology can create strategic vulnerabilities. Some critics point out that this was the result of the Israeli military focusing solely on data processing like an 'AI factory' while ignoring the intuition and experience of human analysts.

Insurmountable Diplomatic Backlash: The AI-driven warfare in the Gaza conflict has sparked global ethical and legal controversies, severely damaging Israel's international legitimacy. Criticism from the International Court of Justice (ICJ), the International Criminal Court (ICC), and international human rights organizations poses a risk of diplomatically isolating Israel.

4 Israeli Military AI Organizations and Companies

Unit 8200 is a core unit under the IDF Intelligence Corps, playing a role comparable to the US NSA and serving as the backbone of Israeli military AI development. Its primary mission is signals intelligence (SIGINT) collection and analysis, processing vast amounts of communication data from various sources into military intelligence in real-time. It operates systems that use machine learning and natural language processing (NLP) to automatically analyze communications in multiple languages—including Arabic, Persian, and English—to identify threats. It also conducts cybersecurity and offensive cyber operations, infiltrating enemy cyber infrastructure to gather information or neutralize systems. One of the unit's most notable achievements is the development of the Lavender (לבנדר) AI system, a machine learning-based tool that synthesizes visual information, mobile metadata, social media activity, and call patterns to assign individuals a probability score from 1 to 100 of being a member of a terrorist organization.

Unit 81 (יחידה 81) is the IDF's technological unit responsible for military R&D and systems integration, focusing on the military application of AI and automation. It develops and operates systems that fuse various sensor data to provide real-time battlefield awareness and researches autonomous operation technologies for unmanned systems such as drones and ground robots. It also develops various tools to support commander decision-making by integrating AI into combat management and command-and-control systems.

The LOTEM Unit (יחידת לוטם), under the C4I Directorate (מפקדת C4I), is responsible for managing and advancing the IDF's IT infrastructure. It oversees the construction and operation of cloud computing, big data analytics, and AI-based data processing systems. LOTEM builds the infrastructure necessary to efficiently store, manage, and analyze massive amounts of military data, ensuring that various AI applications can operate smoothly on this platform.

Unit 3060 (יחידה 3060), within the Military Intelligence Directorate (אגף המודיעין), specializes in operational intelligence and visual information systems, serving as a key organization leading the military integration of AI. It operates systems that use AI to automatically analyze visual data from satellite imagery, drone footage, and ground surveillance cameras to detect enemy movements, weapon deployments, and military facilities in real-time. It also develops and manages systems that synthesize various intelligence sources into comprehensive operational intelligence for commanders.

The Mamram Unit (יחידת ממר"ם) of the Signal Corps manages the IDF's communication networks while developing and operating AI-based communication analysis tools. One of its representative achievements is the Genie (Genie) military chatbot, which uses Retrieval-Augmented Generation (RAG) technology to search for information in vast military databases and provide answers in natural language. Genie supports multiple languages, including Hebrew, Arabic, and English, helping commanders easily access and utilize complex military intelligence.

The Talpiot Program (תלפיות) is the IDF's elite scientific and technological leadership program. It selects a small number of top-tier science students from across the country for nine years of intensive training, grooming them into key personnel for national defense technology. During their service, Talpiot graduates play pivotal roles in developing major Israeli defense systems such as Iron Dome, David's Sling, and the Arrow System. After their discharge, they lead Israel's tech startup ecosystem through the powerful Talpinet (תלפינט) network. They are the primary engine of a virtuous cycle that feeds advanced civilian technology back into the military sector.

Established in 2025 within the DDR&D (מפא"ת), the AI and Autonomy Directorate (מנהלת בינה מלאכותית ואוטונומיה) is responsible for overseeing and coordinating Israeli military AI development at the national level. It integrates and standardizes AI technologies developed by various military units and defense contractors and is responsible for establishing and executing a long-term AI development roadmap. It also researches ways to set ethical standards for AI use and ensure compliance with international legal requirements.

Among private defense contractors, IAI (תעשייה אווירית)—Israel's leading state-owned aerospace company—operates an AI Excellence Center and develops various AI-based military systems. IAI focuses on the development of autonomous drone systems and smart munitions and was the first in the world to commercialize autonomous attack drones like the Harpy (הארפי) and Harop (הרופ). These systems possess a significant level of autonomy, capable of identifying targets and deciding whether to strike independently, placing them at the center of the autonomous lethal weapons controversy.

Rafael Advanced Defense Systems (רפאל) is another major state-owned Israeli defense contractor, famously known as the developer of the Iron Dome missile defense system. The Iron Dome uses AI algorithms to analyze the trajectories of incoming rockets or artillery shells in real-time, selectively intercepting only those projectiles likely to hit civilian areas. AI plays a central role in every step of this process—sensor data processing, threat assessment, interception decision-making, and timing optimization—enabling a high interception success rate of over 90%.

Elbit Systems (אלביט) is a prominent Israeli defense company that holds a leading position in network-centric warfare, autonomous systems, and AI-based object detection technology. Elbit commercialized autonomous attack drone systems by developing the Sky Striker loitering munition and has drastically improved the accuracy of small arms fire through the Smart Shooter AI-based targeting system. It also contributes to crew survivability by developing various AI-based defense systems for tanks and armored vehicles.

Private companies founded by Unit 8200 alumni are vital components of the Israeli defense ecosystem. NSO Group, famous for its Pegasus spyware, developed highly sophisticated cyber-intelligence tools capable of remotely infecting smartphones to collect calls, messages, and location data in real-time. Although embroiled in human rights controversies, this technology is utilized by intelligence agencies worldwide for counter-terrorism and national security purposes.

Wiz is a cloud security startup founded by Unit 8200 alumni. Using AI, Wiz provides solutions that detect and analyze security threats in cloud environments in real-time. It has grown rapidly, reaching a valuation of $12 billion in a short period. The company's technology is used not only by private firms but also for the cloud security of government agencies and military organizations.

Corsight is a facial recognition technology company that develops AI-based biometric identification systems, offering solutions used across security domains including airport security, border control, and military base access management. Kela develops a battlefield AI platform that analyzes real-time combat data to support tactical decision-making. Deep Instinct develops AI-based cybersecurity solutions that detect and neutralize sophisticated cyberattacks.

Palantir is an American company that maintains close ties with Israel and supplies its data analytics platform to the Israeli military. Palantir's Gotham platform is renowned for its ability to integrate and analyze vast quantities of heterogeneous data and extract actionable insights. The Israeli military uses the platform to synthesize data gathered from diverse intelligence sources and to inform operational planning.

Global technology giants including Google, Microsoft, and Amazon also play a significant role in advancing Israel's military AI capabilities, most notably through Project Nimbus. They provide cloud computing infrastructure, AI models, and technical support, supplying the computational resources required for large-scale data processing and machine learning model training. OpenAI likewise supports the military application of natural language processing through its technological cooperation with Israel, contributing to multilingual intelligence analysis and automated translation system development.

The military AI ecosystem that these diverse organizations collectively constitute rests on a foundation of national-level strategic investment, private-sector innovation capacity, and close civil-military collaboration. This ecosystem develops and operates military AI technology that ranks among the most advanced in the world, and it continues to be refined and improved through validation on actual battlefields.

5 Securing Qualitative Military Superiority

Israel's military artificial intelligence strategy is both an inevitable choice for national survival and a modern evolution of its long-standing doctrine of achieving qualitative military superiority. A distinctive civil-military technology ecosystem, sustained national investment, and aggressive battlefield deployment have made Israel a global pioneer in military AI. AI systems such as 'Habsora' and 'Lavender' have fundamentally altered the speed and scale of combat, demonstrating a level of operational efficiency without precedent.

Yet these technological achievements carry a steep cost. The large-scale deployment of AI in the Gaza conflict has generated serious ethical, legal, and diplomatic controversy. The absence of 'meaningful human control,' high permissible civilian casualty thresholds, and the problem of 'accountability gaps' threaten the foundational principles of international humanitarian law and have placed Israel before an international tribunal of opinion. Israel's opposition to new AI regulations on the international stage reads as an attempt to maximize operational freedom through technological advantage, but it simultaneously carries the risks of criticism for violations of international law and diplomatic isolation.

Israel's future depends not merely on sustaining technological superiority, but on establishing an ethical framework for responsible governance of this powerful AI technology and for coexistence within the international community. The paradox of over-reliance on technology and the diplomatic blowback it generates are Israel's strategic vulnerabilities — problems that no better algorithm alone can solve.

Israel's case offers a sharp illustration of both the potential and the darker dimensions of military AI. Other countries, South Korea among them, can draw positive lessons from Israel's experience — among them the value of building an integrated civil-military-academic ecosystem and the importance of battlefield validation — while also developing a deep awareness of the need to establish international legal and ethical standards governing the development and operation of AI weapons systems, and of the importance of 'human-centered design': keeping AI under human oversight rather than pursuing full autonomy. Through Israel's trajectory, the world will gauge the direction warfare takes in the age of AI, and the weight of the responsibility that comes with it.