More than 13 petabytes of Israeli military surveillance data on Palestinians—including biometric records, intercepted communications, and AI-generated targeting lists—sits on Microsoft Azure servers, according to a damning United Nations Human Rights Council report released in late June. The report, authored by Special Rapporteur Francesca Albanese, accuses Microsoft, along with Google and Amazon, of providing infrastructure that is “integral” to an “economy of genocide” in Gaza and the occupied territories. It’s the harshest charge yet leveled against a Big Tech firm for complicity in armed conflict, and it has ignited a firestorm of employee protests, investor pressure, and urgent calls for new regulatory frameworks.

Microsoft’s relationship with Israel runs decades deep. The company operates its largest R&D center outside the U.S. in Israel, and through acquisitions of cybersecurity startups and long-standing government partnerships, it has become woven into virtually every facet of Israeli society—from civilian ministries to the military, police, and the settlement economy. After Hamas’s October 2023 attacks and Israel’s massive military escalation, that partnership intensified dramatically. Israeli military and intelligence agencies’ cloud storage surged nearly 200-fold, exceeding 13.6 petabytes—a volume that dwarfs typical government data usage. Azure now powers real-time data ingestion, translation of intercepted Arabic communications, facial and voice recognition, biometric tagging, and AI tools that analyze information for surveillance and targeting.

The UN Report: Direct Support for Military Operations

The UN Special Rapporteur’s findings are blunt. Microsoft allegedly provided Azure cloud and AI technologies to Israel’s Ministry of Defense (IMOD) under contracts worth at least $133 million. These resources, the report states, “directly support” military planning, real-time intelligence analysis, and—most controversially—algorithmic target selection and automated strike orchestration. A leaked internal Israeli AI system nicknamed “Lavender” is described as using Microsoft’s infrastructure to rapidly sift through thousands of data points and generate actionable target lists with minimal human review, accelerating the pace of airstrikes and ground operations.

The report also highlights Azure’s role in biometric analytics: facial and voice recognition at checkpoints, biometric permit systems, and predictive policing tools that track Palestinian civilian movements. The “sovereign cloud” architecture, designed to keep sensitive data within Israel’s legal jurisdiction, is said to shield these activities from meaningful international oversight, a feature the UN condemns as undermining human rights enforcement.

How the Surveillance Machine Works

Microsoft’s Azure backbone enables what the UN calls “unprecedented” surveillance scale. Biometric, demographic, geospatial, and social media data are ingested in real time, cross-referenced with historic archives, and fed into predictive algorithms that flag individuals as threats. The system’s machine learning models, originally commercial tools, have been repurposed for military intelligence. Intercepted phone calls and messages are automatically translated, transcribed, and analyzed for sentiment and keywords, feeding targets into the “Lavender” AI. Even “air-gapped” classified Israeli systems—normally disconnected from the internet—are reported to run mission-critical workloads on Azure-managed software stacks.

Internal Rebellion: Employees Speak Out

Inside Microsoft, a grassroots movement called “No Azure for Apartheid” has shattered corporate decorum. At major company events, engineers like Joe Lopez and Vaniya Agrawal interrupted CEO keynotes to accuse leadership of complicity in civilian deaths. Microsoft responded by firing several outspoken employees, a move that triggered fresh allegations of censorship. Workers claim emails containing the words “Palestine” or “Gaza” were blocked by internal filters. Similar protests erupted at Google and Amazon, signaling a sector-wide reckoning over the “dual-use” nature of cloud and AI technologies.

Microsoft’s Defense: No Evidence, No Visibility

Microsoft’s official response has been a mix of internal review and careful admission. The company commissioned outside auditors and stated it found “no evidence” its technology was intentionally used to target civilians or violated its Responsible AI Code. But executives conceded a crucial gap: once deployed in sovereign military environments, Microsoft cannot technically or legally observe how its software is used. Critics argue this creates a shield of plausible deniability. “Saying you found no evidence when you have no visibility is meaningless,” said one former employee. The external reviewer remains unnamed, and independent audits are impossible given the operational security.

Profit Amid Conflict

For Microsoft, the controversy sits atop a booming balance sheet. The company’s cloud segment helped drive $70 billion in quarterly revenue and $25.8 billion in net income in fiscal 2025, with government contracts playing a growing role. The Israeli deals are part of a fierce “gold rush” among cloud providers; Amazon and Google compete through projects like the $1.2 billion Nimbus initiative. Microsoft’s contracts allowed decentralized purchasing by multiple Israeli military branches, police, and intelligence units, deepening its entrenchment and, critics say, its legal exposure.

The Human Toll

The Gaza Health Ministry reports more than 50,000 Palestinians killed by April 2024, with entire families erased in airstrikes assisted by AI targeting. The UN report, buttressed by testimony from rights groups and tech workers, contends that cloud-based automation has intensified the precision, speed, and scale of violence to a level some legal scholars argue meets the threshold for genocide, though no binding court ruling has been made. Civilian casualties, the report notes, are not collateral damage but a direct result of mass profiling and automated strike recommendations that endanger humanitarian workers and medical personnel.

Microsoft’s legal exposure stretches beyond public relations. With Azure servers located in Europe and subject to International Court of Justice statutes, the company could face liability if Israeli actions are definitively classified as war crimes or genocide. Activist shareholders have filed resolutions demanding greater transparency, and the Boycott, Divestment, Sanctions (BDS) movement has added Microsoft to its official boycott list. Regulatory investigations loom, and the European Union is considering stricter rules for cloud providers operating in conflict zones.

The Bigger Picture: Tech’s Neutrality Myth Shattered

The Microsoft-Israel storm is only the most acute example of a paradigm shift. For decades, tech giants sold their platforms as neutral utilities. Employee activism, consumer pressure, and global scrutiny have obliterated that myth. “This didn’t start with a broad policy debate; it started with engineers refusing to build tools for oppression,” said a spokesperson for the No Azure for Apartheid group. The industry must now confront uncomfortable questions: Can companies meaningfully prevent weaponization of their platforms? Should shareholders or governments set red lines? The era of hands-off ethics is over.

Microsoft’s predicament serves as a warning. The company’s next moves—whether it tightens contract oversight, withdraws from certain sectors, or doubles down on its sovereign cloud strategy—will be watched closely by activists, investors, and global regulators. The world is learning that the code powering the cloud is never just code. It can become a weapon, and the companies that write it are being forced to reckon with the consequences.