A joint investigation by The Guardian, +972 Magazine, and Local Call has exposed how Israel’s elite military intelligence unit, Unit 8200, harnesses Microsoft Azure to store and process intercepted Palestinian phone calls on a staggering scale—up to a million calls an hour. Operational since 2022, this cloud infrastructure now underpins a suite of AI-driven surveillance and targeting algorithms that accelerate airstrikes, detentions, and blackmail operations across Gaza and the West Bank.

The findings, corroborated by United Nations reports and whistleblower testimony, place Microsoft at the heart of a digital warfare apparatus that critics say has weaponized commercial cloud technology. Internal documents and interviews with 11 sources from the company and Israeli military intelligence detail how Azure’s segregated “sovereign” environment became the backbone of a mass surveillance system, raising explosive questions about corporate complicity, human rights, and the accountability of big tech in armed conflict.

A Cloud Engineered for Mass Interception

The collaboration traces back to a pivotal 2021 meeting between Unit 8200 commander Yossi Sariel and Microsoft CEO Satya Nadella. Shortly after, a custom Azure environment was carved out exclusively for “sensitive workloads,” with Microsoft employees working alongside Israeli military contractors on a near-daily basis to design a security architecture that could ingest and analyze colossal volumes of intercepted audio. By July 2025, this data hoard had swollen to 11,500 terabytes—equivalent to roughly 200 million hours of voice recordings—hosted in Azure data centers in the Netherlands and Ireland.

The operational tempo accelerated dramatically after Hamas’s October 2023 attack. Israeli military data stored on Microsoft infrastructure ballooned nearly 200-fold, surpassing 13.6 petabytes. This exponential growth reflected not just the scale of interception but the integration of advanced AI tools. Real-time translation of Arabic, voice and facial recognition at checkpoints, predictive analytics to identify adversaries, and automated target selection—via platforms like “Lavender”—all draw on Azure’s processing power. Sources say the system can flag potential targets with minimal human review, collapsing the decision cycle for lethal strikes.

A landmark UN Human Rights Council report authored by Special Rapporteur Francesca Albanese has thrust these practices into the international legal spotlight. Albanese’s findings, supported by human rights groups and Israeli whistleblowers, accuse Microsoft, Google, and Amazon of “profiting from the Gaza genocide.” The report specifically highlights how infrastructure sold under standard commercial contracts enables mass surveillance that is disproportionate, indiscriminate, and inextricably linked to civilian harm.

Microsoft denies any direct knowledge of the intercepted call data or of consulting on surveillance systems. An internal review, the company claims, uncovered “no evidence” that Azure or its AI tools were used to directly harm civilians. Yet Microsoft concedes a fatal limitation: once technology is placed within a government’s sovereign cloud, the company has virtually no visibility into its downstream applications. This accountability gap, critics charge, creates a veneer of deniability while the platform functions as a force multiplier for military operations that international jurists argue may violate the Geneva Conventions and the Rome Statute of the International Criminal Court.

Employee Revolt and Industry Reckoning

Inside Microsoft, the crisis has ignited a firestorm. The activist group “No Azure for Apartheid,” composed of current and former employees, has led public protests, disrupted the company’s 50th anniversary event, and staged walkouts at the Build developer conference. Their demands: an immediate halt to all military contracts with Israel and consistent application of Microsoft’s own human rights principles. Management has responded by firing several vocal employees, deepening the rift and fueling a wider culture of dissent.

Microsoft is not alone. “No Tech for Apartheid” campaigns at Google and Amazon have mounted similar pressure over the $1.2 billion Project Nimbus cloud contract with the Israeli government. Shareholders and institutional investors now demand transparent human rights impact assessments, and legislation in the U.S. and Europe seeks to clamp down on the military misuse of dual-use technologies. The outcry reflects a tectonic shift: a new generation of tech workers expects their employers to adhere to ethical norms, and violations risk talent flight, reputational ruin, and litigation.

The Double-Edged Sword of Sovereign Cloud

Azure’s very design illustrates the profound dilemma. Its sovereign cloud configuration—engineered for government-level data isolation and legal control—is a prized feature for nation-states seeking digital independence. Yet those same features shield processes from external audit, enabling mass surveillance and automated targeting without accountability. Azure’s scalability, AI toolset, and rapid deployment capabilities have made it a dominant choice for defense ministries, but they also render it an irresistible instrument for waging data-driven warfare.

The UN investigation underscores that the erosion of oversight is not accidental. When commercial contracts carry only vague “acceptable use” clauses and the provider cannot inspect customer payloads, the distance between a platform’s marketed purpose and its battlefield application vanishes. For Microsoft, the challenge is existential: how to reconcile the immense commercial pressures of the cloud market with emergent legal duties under international humanitarian law.

Conclusion: The Unfinished Search for Governance

The leaked documents and parallel UN findings mark a turning point. They lay bare a system in which cloud computing, artificial intelligence, and military ambition fuse seamlessly, leaving tech companies enmeshed in conduct that human rights bodies label as crimes. Microsoft’s Azure has not simply been rented out; it has become an integral component of a surveillance-and-strike architecture that leaves civilians in Gaza with no refuge from the digital gaze.

Without robust, enforceable governance—both inside the corporation and through multilateral treaties—the militarization of everyday technology will only deepen. As the International Criminal Court and national regulators weigh their next steps, the immediate onus falls on Microsoft’s leadership to decide whether commercial success can coexist with the imperative to prevent atrocity. The world is watching, and the ethical boundaries of technology hang in the balance.