How US Tech Giants Supplied AI Models to Israel and Its Impact on Warfare

TEL AVIV, Israel — U.S. tech companies have significantly aided Israel in its military operations against alleged militants in Gaza and Lebanon by providing advanced artificial intelligence (AI) and computing services. This acceleration in AI usage has prompted serious concerns about the collateral damage and rising civilian deaths resulting from these technologies.

For years, military organizations globally have collaborated with private firms to create specialized autonomous weapons. However, the recent conflicts involving Israel showcase a notable instance where commercial AI developed in the U.S. has been employed directly in active combat—despite ethical issues regarding their original purpose.

The Israeli military extensively utilizes AI to analyze large amounts of intelligence data, intercepted communications, and surveillance footage to identify suspicious behaviors and track its adversaries' activities. Following the unexpected Hamas attack on October 7, 2023, it has been reported that the military's use of Microsoft and OpenAI technologies dramatically increased.

An investigation revealed that the implementation of these AI systems for targeting and decision-making raised concerns over errors stemming from inaccurate data or biased algorithms. This investigation was based on internal documents and interviews with officials and employees within the industry.

After the headline-grabbing assault that resulted in approximately 1,200 fatalities and over 250 hostages, Israel's strategy focused on the total dismantling of Hamas, leading military leaders to describe AI as a "game changer" in rapidly identifying targets. Since the inception of warfare, more than 50,000 individuals have lost their lives in Gaza and Lebanon, with nearly 70% of buildings in Gaza heavily damaged, according to local health data.

Heidy Khlaaf, chief AI scientist at the AI Now Institute, emphasized the ramifications of this development: "This is the first confirmation we have gotten that commercial AI models are directly being used in warfare... the implications are enormous for the role of tech in enabling this type of unethical and unlawful warfare going forward."

Significant Rise in AI and Cloud Computing Usage

Among U.S. tech companies, Microsoft has long-standing ties with the Israeli military, which intensified post-Hamas assault. These relationships stressed Israel's own computing resources, pushing them to rely on external vendors, as highlighted by Col. Racheli Dembinsky, the military's chief IT officer. During a recent briefing, she pointed out that AI had given Israel "very significant operational effectiveness," showcasing logos from Microsoft Azure, Google Cloud, and Amazon Web Services in the background.

According to the investigation, the Israeli military's use of Microsoft and OpenAI's AI surged almost 200 times higher in the days following the October 7 attack. The data stored on Microsoft servers doubled from that point until July 2024, amassing over 13.6 petabytes—that roughly equates to 350 times the required digital memory to store every book in the Library of Congress. Moreover, deployment of Microsoft’s extensive server banks by the military increased by almost two-thirds within the initial two months of conflict.

Microsoft has chosen not to comment on the relationship with the Israeli military, despite inquiries. However, the company asserts that it values human rights and aims to promote a responsible technology agenda. In its lengthy Responsible AI Transparency Report for 2024, it commits to managing potential generative AI risks throughout development but does not elaborate on military contracts.

Advanced AI models are supplied by OpenAI, known for ChatGPT, primarily through Microsoft's Azure cloud services, which the Israeli military has been acquiring. Despite Microsoft’s backing, OpenAI states that it doesn’t partner with the Israeli military, having previously updated its usage policies to permit "national security use cases that align with our mission."

When questioned, the Israeli military refrained from providing detailed comments regarding the integration of American commercial AI and asserted that its analysts use AI to independently verify target intelligence alongside senior officers as required by international law, balancing military goals and potential civilian harm.

Partnerships with Other Technology Firms

In addition to Microsoft, Google and Amazon are also suppliers of cloud computing and AI capabilities under a $1.2 billion contract known as Project Nimbus, inked in 2021, allowing Israel to trial its homegrown AI targeting systems. The Israeli Defense Forces (IDF) also utilize server infrastructures from Cisco and Dell. Furthermore, Palantir Technologies partners with the Israeli military, supplying AI solutions intended to support military operations.

Following OpenAI's policy adjustments to support national security purposes last year, Google similarly revised its ethical guidelines to remove previous restrictions against using its AI for military applications.

Google remarked that its commitment remains towards responsible development and deployment of AI, which contributes positively to global security.

The Functionality and Reliability of AI in Warfare

The Israeli military exploits Microsoft Azure to aggregate data received from widespread surveillance, which gets transcribed and translated. This includes various forms of communication like voice calls and text messages. An officer tied to these systems described relying on Azure for swift searches within massive text documents, enhancing the overall target identification process.

Despite the advancements, significant errors can occur due to multiple factors associated with AI. Intelligence gathered via phone interceptions must be verified through original audio or accurately translated manuscripts, hence requiring skilled personnel fluent in Arabic to assess translations. Nevertheless, reports indicate that inaccuracies in machine translations from Arabic to Hebrew have previously led to devastating targeting errors.

AI, military, Israel, technology, surveillance