Tag: Artificial Intelligence

  • Maven in the Iran War

    Maven in the Iran War

    In late February 2026, the United States and the Zionist regime launched coordinated military operations under the names “Epic Fury” and “Roaring Lion” against Iran—operations considered the largest U.S. military action since the 2003 invasion of Iraq. During this conflict, CENTCOM struck over 11,000 targets, while Iran responded with thousands of missiles and drones. Iran’s Supreme Leader, Ayatollah Khamenei, was martyred in the initial attacks. However, these statistics alone do not capture the historical significance of this war; rather, what distinguishes this conflict is the central role of artificial intelligence, autonomous systems, and commercial technologies—which no longer serve merely as support tools but have been placed at the core of operations.

    The “Maven” project had been initiated back in 2018. Prior to the emergence of large language models, this unit, in collaboration with several companies, was supporting the Maven project to advance “computer vision”—an AI capability that enables the detection and differentiation of objects in satellite imagery, assisting analysts in reviewing visual data.

    The Maven project helped accelerate the deployment of machine learning in the fields of reconnaissance, surveillance, intelligence analysis, and geospatial intelligence. This legacy ultimately led to the development of the “Maven Smart System” by Palantir Technologies, which today has become the cornerstone of AI-driven U.S. military operations.

    Background

    The U.S. Department of Defense has designated Palantir as the “sole responsible supplier” for the Maven AI system—a system used for field data analysis, target identification, and military decision support. The author emphasizes that this contract marks the transition of the Maven project from the experimental phase to a permanent U.S. military infrastructure level. According to the sources cited, the system is capable of analyzing satellite data, drone footage, sensor inputs, and field intelligence in real time, generating a “unified operational picture” for military commanders. A notable point is the direct link between Palantir and America’s broader strategy in the domain of “AI-enabled warfare,” which itself indicates that the U.S. government is gradually outsourcing its intelligence analysis and targeting infrastructure to private Silicon Valley companies—a trend that, in his view, blurs the boundaries between government, military, and tech firms.

    Analysis and Findings

    A. The Role of Silicon Valley in the U.S. War Department

    Silicon Valley’s role in the U.S. Department of Defense has expanded significantly in recent years; this collaboration has moved beyond limited research projects in cybersecurity over the past decade and has now become a central pillar of the U.S. military’s digital transformation. Recognizing the rapid pace of emerging technologies such as artificial intelligence, machine learning, and big data analytics, the Pentagon has increasingly turned to partnerships with Silicon Valley tech companies to enhance decision-making, combat effectiveness, and intelligence superiority. A prime example of this trend is the massive contracts with firms like Palantir Technologies, Google Cloud, and Microsoft, which are involved in projects such as Maven and the Joint Warfighting Cloud Capability. The U.S. Department of Defense now has over 200 active contracts with companies based in the San Francisco Bay Area, a large portion of which are directly linked to digital warfare operations. This extensive partnership, however, has not always been without challenges. Today, the bulk of AI system development for the Department of Defense is managed by private contractors based in Silicon Valley. From the Pentagon’s perspective, collaboration with Silicon Valley is a strategic necessity. The engineers and researchers in this region are among the world’s foremost experts in areas such as image processing, data analysis, and deep learning platform design. By establishing the Joint Artificial Intelligence Center (JAIC) in 2018, the Pentagon effectively opened a gateway for direct engagement with tech companies. This body is tasked with adapting emerging civilian technologies for military applications and setting ethical and technical standards for the use of AI in warfare. Through this model, the U.S. military can innovate at a pace close to the private sector, without being bogged down for years in bureaucratic procedures. Silicon Valley has become a vital component of America’s modern warfare infrastructure, where the line between commercial technology and military application is increasingly blurred. Companies like Palantir, Anduril Industries, and SpaceX now play pivotal roles in combat data analysis, autonomous drones, and military satellite communications. By September 2024, the share of commercial technology firms in the U.S. Department of Defense’s research budget had risen from 11% in 2015 to over 37% in 2024. This extensive partnership, while enhancing military efficiency, has also raised new concerns about transparency, social oversight, and the over-reliance of national security on the private sector and tech investors.

    Image of part of the terms of the US Department of Defense contract with Palantir

    B. The Role of Maven in the Attack on Iran

    The Maven system combines satellite imagery, drone videos, radar data, and signals intelligence into a unified interface, allowing operators to identify targets, recommend weapon types, and generate attack plans almost simultaneously. The results have been remarkable: over a thousand targets were struck in the first 24 hours of operations—a pace unimaginable if relying solely on human analysis. This process has continued using only 10% of the analysts that would previously have been required for such a level of operations.

    However, the system’s limitations are equally notable. Maven’s overall accuracy is estimated at around 60%, compared to roughly 84% for human analysts. Despite this, Palantir’s Chief Technology Officer has described it as “the first large-scale AI-enabled combat operation”—a characterization that raises serious questions about the ethical considerations of AI-based targeting and the adequacy of civilian protection mechanisms. These very concerns about protective safeguards have also led to the recent controversy between Anthropic and the Department of Defense.

    C. The History of Maven

    Project Maven is one of the U.S. Department of Defense’s programs in the field of applying artificial intelligence to military data analysis, launched in 2017. Its official name was the “Algorithmic Warfare Cross-Functional Team,” and its primary objective was to use machine learning for the automated analysis of images and videos from reconnaissance drones. The project was initially designed to accelerate the processing of vast battlefield data, particularly in Middle Eastern operations. Early sources about the program’s inception were published in Department of Defense statements in 2017. Project Maven focused specifically on drone imagery analysis—using machine vision algorithms to detect individuals, vehicles, and suspicious objects in aerial footage. Previously, such imagery had to be analyzed manually by human personnel, a highly time-consuming process. The project’s goal was to reduce the workload on analysts and increase the speed of operational decision-making. Pentagon official reports emphasized that the system played a “decision-support” role and was not a complete replacement for human personnel. The name Palantir Technologies has frequently been associated with U.S. defense data analysis projects. Palantir is a major government contractor in the fields of big data analytics and intelligence software, with its products such as “Palantir Gotham” used in defense and security agencies. Although Project Maven was broadly executed with the participation of multiple contractors, media attention was largely focused on Google’s initial partnership with the Pentagon, while companies like Palantir were also active in the defense technology ecosystem related to data analysis and AI.

    According to statistics released by the White House on April 8, the U.S. military struck over 13,000 targets in Iran during the first 38 days of the war—including more than 2,000 command-and-control centers, 1,500 air defense targets, and 1,450 targets related to industrial infrastructure.

    The White House emphasized that “all these targets were legitimate and military in nature.” However, the New York Times and other media outlets, through reports and images, have documented the destruction of civilian facilities. These reports have sparked a wave of criticism and raised the question of whether over-reliance on AI systems has led to targeting errors and increased civilian casualties.

    During operations against Iran, the Maven Smart System—a Palantir product—played a central role in gathering and analyzing battlefield intelligence. According to Cameron Stanley, the Pentagon’s senior official for digital and artificial intelligence, the system collected data on enemy force positions from various sources, including radar signals, satellite and drone imagery, electronic communications, and other intelligence feeds, combining them all into a unified operational picture of the battlefield.

    Stanley, who presented these remarks at Palantir’s “AIPCon 9” conference on March 12, stated that Maven is not limited to target identification. According to him, the system can also determine the most suitable friendly unit to engage each target and provide U.S. combat units with specific operational options—including attack routes, required munitions types, and other relevant information for mission execution. He explained that previously, commanders had to consult eight or nine different systems simultaneously to make battlefield decisions, but Maven now aggregates all this information into a single visual interface. In this way, a commander can identify a target, select the appropriate attack package, and initiate the operational process with just a few clicks.

  • Endless Cyber Conflicts: From Tehran to Tel Aviv

    Endless Cyber Conflicts: From Tehran to Tel Aviv

    The Twelve‑Day Cyber War
    The twelve‑day conflict between Iran and Israel, which took place from June 11 to 17, marked a turning point in the history of Middle Eastern cyber warfare. Initiated by Israel’s Operation “Rising Lion”, the confrontation saw a 700 percent increase in cyberattacks against Israeli infrastructure. Unit 8200, the largest military unit within Israel’s Defense Forces and often compared to the U.S. NSA or the UK’s GCHQ, played a pivotal role. Responsible for clandestine operations, SIGINT collection, codebreaking, counter‑espionage, cyber warfare, and surveillance, Unit 8200 has become one of the most sophisticated cyber units globally. Its importance is underscored by the fact that its veterans founded companies such as Check Point, CyberArk, Claroty, CyCognito, and Palo Alto Networks, which both support national security and serve as instruments of Israel’s cyber soft power abroad.

    Unit 8200’s Role in the Twelve‑Day War
    During the June 2025 conflict, Unit 8200 carried out a wide range of cyber operations against Iran. According to CloudSEK, between June 11 and 17, more than 35 pro‑Iranian hacker groups launched coordinated cyberattacks on Israeli military, governmental, and infrastructure targets. In response, Israel activated only four to five groups, but the sophistication of Unit 8200 and affiliated groups such as Predatory Sparrow far outweighed sheer numbers. This group, regarded as the successor to INDRA, employed unique cyberattack methods against Iran’s critical infrastructure, having previously inflicted significant damage on rail systems, steel plants, and fuel stations. Israel’s National Cyber Directorate (INCD) reported that in 2024, the number of alerts doubled, with cyberattack damages estimated at over $3 billion. These figures highlight the unprecedented intensity of the conflict, which went far beyond simple DDoS attacks or website defacements, encompassing complex intrusions into critical infrastructure, theft of sensitive data, and disruption of industrial systems.

    Technological Transformation and AI Tools
    One of Unit 8200’s most significant advancements has been the development and deployment of advanced artificial intelligence systems. According to The Guardian and +972 Magazine, Unit 8200 is building a large language model (LLM) similar to ChatGPT, trained on millions of intercepted Arabic conversations from Palestinians. Still in its training phase, this AI tool is designed to analyze, translate, predict, and summarize text, dramatically enhancing the unit’s surveillance and intelligence capabilities. INCD’s new strategy for 2025–2028 includes the creation of a “Cyber Dome”, an AI‑based national cyber defense system aimed at countering growing threats from Iran and Hamas. At a conference in Tel Aviv, Unit 8200’s commander revealed that AI technology had already been used to select Hamas targets. Reports also indicate that Microsoft and other major U.S. tech firms have supported Unit 8200 by providing Azure cloud services, enabling the migration of 70 percent of sensitive data to the cloud. This reflects the deep integration of Western technologies into Israel’s cyber warfare structure.

    Analysis of Tactics and Operational Methods
    In the twelve‑day war, Unit 8200 employed a wide array of advanced tactics, showcasing significant evolution in cyber warfare methods. According to Dark Reading, the attacks against Iran unfolded in three distinct phases: the first marked by massive DDoS waves and initial intrusion attempts; the second focused on targeting managed service providers (MSPs); and the third, ongoing phase, characterized by slower but far more sophisticated operations.

    Iranian attackers, who previously relied on outdated malware and easily blocked scripts, have shifted to using remote management tools (RMMs) and legitimate Windows software, making detection and blocking far more difficult. Exploitation of zero‑day vulnerabilities, which once took days or weeks, now occurs within 30–40 minutes.

    This dramatic increase in speed and efficiency reflects growing cooperation among Iran‑aligned groups, with shared intelligence, infrastructure, and R&D capabilities. In response, Unit 8200 has deployed advanced tools to target high‑value and unconventional assets.

    Strategic Consequences and Regional Impact
    The twelve‑day cyber war had profound strategic implications for the Middle East. According to the Lieber Institute, in the six months following the Gaza war’s outbreak in October 2023, about 60 percent of Iran’s cyber operations were directed against Israel, intensifying across all fronts from espionage to infrastructure attacks. This escalation was reinforced by Tehran’s allies in Lebanon, Iraq, and Syria. Israel’s National Cyber Directorate, which issued 367 alerts in 2023, raised the figure to 736 in 2024, including 518 high‑priority “red alerts.”

    With much of Iran’s defensive infrastructure destroyed and several senior military officials killed by Tel Aviv’s air campaign, Tehran’s strategic options have been significantly constrained. This operational damage reduces the feasibility of a major conventional response in the short term, making asymmetric tools—especially cyber warfare—far more attractive. Meanwhile, Saudi Arabia and the UAE, scarred by the Shamoon attack that wiped 30,000 computers and temporarily crippled Saudi Aramco, are increasingly turning to cyber capabilities as instruments of national security. Western states, on the other hand, continue to portray Iran as a disruptive actor in this domain, though experience shows Israel does not wait passively—conducting periodic attacks reminiscent of the Stuxnet virus.

    Below, three possible scenarios for the evolution of this new type of warfare are outlined:

    Scenario One: AI‑Driven Cyber War
    In the first scenario, Unit 8200 leverages its advances in artificial intelligence to enter a new phase of cyber warfare. Once its developing LLM is completed and operational, the unit will be able to process vast amounts of surveillance data in real time. This capability will enable faster target identification, more accurate enemy movement predictions, and automated cyber operations. The planned AI‑based Cyber Dome (2025–2028) could evolve into a multilayered defense system that not only repels attacks but also acts preemptively against potential threats. In this scenario, Unit 8200 is expected to use deep learning algorithms to detect complex patterns in network traffic and anticipate zero‑day attacks before they occur. Deeper collaboration with U.S. tech giants such as Microsoft, Google, and Meta would provide massive computing resources and training data, potentially leading to the development of autonomous cyber weapons capable of independently deciding on targets and attack methods. In response, Iran is likely to invest heavily in AI‑based defensive capabilities and may seek cooperation with countries such as Russia and China to access advanced technologies.

    Scenario Two: Expansion of Cyber Warfare to Regional Critical Infrastructure
    In this scenario, cyber conflict extends beyond Iran and Israel to encompass the entire Middle East. Unit 8200 may adopt a strategy of “maximum cyber pressure,” targeting not only Iran’s military and nuclear infrastructure but also vital civilian systems such as electricity grids, water networks, transportation, and banking. Reports indicate that the CyberAvengers group, affiliated with Iran’s Revolutionary Guard, targeted Unitronics PLC devices across multiple U.S. critical infrastructure sectors—including water and wastewater systems—for a year beginning October 7. Such attacks highlight the growing tendency to strike critical infrastructure, with potentially catastrophic consequences for civilian populations. In this scenario, Unit 8200 would likely exploit its capabilities to penetrate industrial control systems (ICS) and SCADA networks in Iran, aiming to cause widespread disruptions that could trigger social unrest. Conversely, Iranian cyber groups may attempt to strike Israel’s critical infrastructure, including water treatment facilities, power plants, and transportation networks. This escalation could evolve into a “cyber cold war”, where both sides continuously probe each other’s defenses and search for vulnerabilities without necessarily launching full-scale destructive attacks.

    Scenario Three: Convergence of Cyber Warfare with Physical Operations
    The third scenario is the most complex and dangerous, involving the integration of cyber operations with physical attacks and psychological warfare. Here, cyberattacks are not used as standalone tools but as part of a multi-layered strategy. For example, Unit 8200 might first disable Iran’s air defense systems through cyber intrusions, then conduct airstrikes, followed by information operations to amplify psychological impact. Reports suggest that during recent operations, false messages attributed to the Israeli military—warning of bomb shelter attacks—were disseminated by pro-Iranian groups to spread panic among civilians. Such hybrid operations are highly destructive, as defending against them requires coordination across military, security, and civilian sectors. Leveraging its AI capabilities, Unit 8200 could design sophisticated disinformation campaigns, using deepfakes and AI-generated content to manipulate public opinion. In this scenario, the boundaries between cyber warfare, information warfare, and conventional conflict blur entirely. Iran, in turn, may coordinate cyber-physical attacks through its regional allies against Israeli and allied interests. Israel itself has long engaged in such operations, often playing a role in initiating or facilitating unrest inside Iran.

    Conclusion: An Uncertain Future in Cyberspace
    The twelve-day conflict in June demonstrated that a new era of cyber warfare has begun—an era in which Israel’s Unit 8200, armed with advanced AI technologies, extensive international cooperation, and innovative tactics, seeks to maintain cyber superiority in the region. Yet, as reports indicate, this advantage is fragile, with Iran rapidly expanding its own capabilities. The three scenarios analyzed—AI-driven escalation, expansion to critical infrastructure, and convergence with physical operations—all point to the potential for dangerous intensification in the future. What is clear is that the cyber war between Iran and Israel has entered a new phase, one in which international laws and norms are repeatedly violated and civilian populations bear the greatest burden. The development of Israel’s Cyber Dome and Iran’s likely countermeasures could push the region into a cyber arms race with no foreseeable end. Ultimately, the international community must remain alert to the growing risks of this cyber conflict and work to establish legal and diplomatic frameworks to contain it before its consequences spiral out of control.

  • Amazon’s Cloud Services Linked to Israeli Military Operations in Gaza

    Amazon’s Cloud Services Linked to Israeli Military Operations in Gaza

    According to the report, Israeli defense companies—particularly Rafael Advanced Defense Systems and Israel Aerospace Industries—have directly utilized Amazon Web Services (AWS) to support their technological infrastructure. These services played a role in the operational framework of the Israeli military during its campaigns in Gaza.

    The exposure of Amazon’s collaboration with Israel’s military underscores a growing trend in the digital age: the blurring of lines between civilian tech industries and military applications. Companies that appear to operate in commercial and technological sectors are increasingly becoming active participants in armed conflicts.

    Cloud computing, battlefield data analytics, and artificial intelligence have become integral to both offensive and defensive military strategies. In this context, Amazon has transitioned from a service provider to a direct enabler of Tel Aviv’s war machine.

  • Digital Assassination: AI-Enabled Targeting of Hamas Commanders

    Digital Assassination: AI-Enabled Targeting of Hamas Commanders

    According to four Israeli officials, following the October 7 attack on the occupied territories—which resulted in the death of over 1,200 Israelis and the taking of 250 hostages—artificial intelligence technologies were rapidly approved for field use. According to them, this led to cooperation between Unit 8200 and reserve forces in the “Studio” to swiftly develop new AI capabilities.

    Over the past 18 months, the occupying regime has combined artificial intelligence with facial recognition software to match partially covered or injured faces with real identities. It has also used AI to compile lists of potential airstrike targets and developed an Arabic-language AI model to power a chatbot capable of scanning and analyzing text messages, social media posts, and other Arabic-language data.

    In late 2023, the Zionist regime was attempting to assassinate Ibrahim Biari, the commander of Hamas’s Jabalia Central Battalion, who was allegedly involved in planning the October 7 attack. The regime’s military intelligence intercepted Biari’s communications with other Hamas members but was unable to pinpoint his exact location. Therefore, they turned to an AI-powered audio tool—one that analyzed various sounds such as sonic booms and airstrikes.
    After the approximate location of Ibrahim Biari’s communications was identified, Zionist military officials were warned that the area—which included several apartment complexes—was highly populated. They stated that to ensure Biari’s death, multiple buildings needed to be targeted in the airstrike. The operation was approved and carried out.

    Since then, the regime’s military intelligence has also used this audio tool—alongside maps and images of Gaza’s complex underground tunnels—to search for hostages. According to two Israeli officers, the tool has improved over time to identify individuals with greater accuracy.

    Tracking the location of individuals
    According to three informed Israeli and American officials (whose names were not disclosed in the report), Israeli officers turned to a new military technology integrated with artificial intelligence to track their targets. This technology had been developed about a decade ago but had not yet been used on the battlefield. The effort to locate Martyr Ibrahim Biari provided fresh motivation to enhance the tool. Engineers from Unit 8200—the occupying regime’s equivalent of the U.S. National Security Agency—rapidly integrated AI into the system.

    Shortly afterward, the regime’s security personnel intercepted Mr. Biari’s phone calls and tested the AI-based audio tool, which identified the approximate location of his communications. Using this information, the Zionist regime ordered an airstrike on the area on October 31, 2023, resulting in the martyrdom of Ibrahim Biari.

    According to a report by Airwars, a London-based conflict monitoring organization, due to the approximate nature of the AI-generated location, more than 125 civilians were also killed in the strike! Based on interviews with nine American and Israeli defense officials, the audio tool is just one example that demonstrates how the regime has used the Gaza war to test and rapidly deploy AI-supported military technologies on an unprecedented scale.

    Behind-the-Scenes Companies and Reactions to the Committed Atrocities
    Many of these efforts resulted from collaboration between active-duty soldiers in Unit 8200 and reservists working at tech companies such as Google, Microsoft, and Meta. According to these individuals, Unit 8200 has established a central hub called the “Studio,” which serves as an innovation center and a place to connect experts with AI projects. Reservists from the aforementioned major companies are involved in this unit. The Studio’s primary mission is to rapidly adapt artificial intelligence technologies to meet military needs.

    While the regime rapidly continued to develop its artificial intelligence arsenal, the deployment of these technologies has led to mistaken identifications, wrongful arrests, and civilian deaths. According to European and American defense officials, no other power has been as active as the occupying regime in testing AI tools in real battles. This situation provides an initial glimpse into how such technologies might be used in future wars, as well as a warning of how these technologies could malfunction or be misused, resulting in the killing of civilians.

    Meta and Microsoft declined to comment on the killings carried out in the Gaza war using artificial intelligence, but Google, in an effort to clear itself of accusations, stated: “We have employees who serve as reservists for the regime in various countries around the world. The work these employees do as reservists is not related to Google.”

    Avi Hasson, CEO of the nonprofit organization Startup Nation Central, which connects investors to companies in the occupied territories, said, “Reservists from Meta, Google, and Microsoft have played a vital role in advancing innovation in drones and data integration.” He added, “Reservists brought specialized knowledge and access to key technologies that were not available in the military.”

    Avi Hasson

    Artificial Intelligence in the Drone Manufacturing Industry
    The Israeli regime’s military soon began using artificial intelligence to upgrade its drone fleet. Aviv Shapira, founder and CEO of XTEND—a software and drone company collaborating with the regime’s military—said that AI-based algorithms have been used to develop drones capable of locking onto and tracking targets from a distance. “In the past, targeting relied on focusing on the image of the target. But now, AI can identify and track the object itself—whether it’s a moving vehicle or a person—with very high accuracy,” he said. Shapira added that his main clients—the occupying regime’s military and the U.S. Department of Defense—are aware of the ethical implications of using AI in warfare and are engaging in discussions about the responsible use of this technology.

    Aviv Shapira

    The Arabic Language: A Battleground for the Enemy to Influence Public Opinion
    According to three Israeli officers, one of the tools developed by the “Studio” was an Arabic-language artificial intelligence model, classified as a large language model. (This LLM was previously reported by the news site +972.) Developers had previously faced difficulties in creating such a model due to the lack of sufficient Arabic-language data to train the technology. Even when such data was available, it was mostly in standard written Arabic, which is far more formal than the dozens of different dialects used in spoken Arabic.

    According to Israeli officers, the occupying regime’s military did not face this problem. The illegitimate government had access to decades’ worth of intercepted text messages, recorded phone conversations, and social media posts collected in colloquial Arabic. As a result, in the first few months of the Gaza war, Zionist elements developed the large language model and designed a chatbot capable of handling queries in Arabic. They integrated this tool with multimedia databases, enabling Zionist analysts to conduct complex searches across images and videos.

    When the Zionist regime assassinated Sayyed Hassan Nasrallah in September, the developed chatbot analyzed reactions to his martyrdom across the Arabic-speaking world. The technology distinguished between different dialects within Lebanon to assess public sentiment and helped the regime determine whether there was pressure from public opinion to carry out a retaliatory strike.

    It is also reported that, at times, the chatbot was unable to recognize certain modern slang terms and words that had been phonetically transliterated from English into Arabic. One of the officers stated that this required Zionist intelligence officers specialized in various dialects to review and correct the results.

    The chatbot also occasionally produced incorrect results. For example, instead of returning an image of a weapon, it would display a picture of a pipe. Nevertheless, they stated that the AI tool significantly accelerated the process of research and analysis.

    Facial Recognition and Identification of Individuals
    At temporary checkpoints set up between northern and southern Gaza following the October 7 attacks, the regime also deployed cameras capable of capturing high-resolution images of Palestinians and sending them to an AI-based facial recognition program. According to two intelligence officers from the regime, the system sometimes struggled to identify individuals whose faces were covered. This led to the detention and interrogation of Palestinians who had been mistakenly identified by the facial recognition system.

    The regime also used artificial intelligence to process and refine data collected by intelligence officials about Hamas members. Prior to the war, the regime had developed a machine learning algorithm called “Lavender,” designed to rapidly sort through data and assist in identifying low-level militants. This algorithm was trained on a database of confirmed Hamas members and aimed to predict who else might be affiliated with the group.

    According to reports, “Lavender” helped the occupying army compile a list of 37,000 human targets based on their association with Hamas. Although the system’s predictions were not flawless, the regime used it at the beginning of the Gaza war to assist in selecting targets for attack!