Microsoft has concluded the external investigation conducted in response to concerns related to how the Israeli Ministry of Defense (IMOD) was using its cloud platform.

In a summary document, Microsoft said that this is the "final update" regarding the matter, and lists some follow-up steps it is taking to "enhance governance and oversight going forward."

microsoft logo
– Getty Images

The investigation in question was launched in August 2023, following reports from The Guardian that Israel's military surveillance agency, Unit 8200, was using Microsoft Azure to store audio files from millions of phone calls made by Palestinians and had even been used to coordinate military attacks on Palestinians.

The following month, Microsoft cut off some of the services it was providing to the Israeli Ministry of Defense (IMOD) following an internal review.

At the time, Microsoft's vice chair and president, Brad Smith, noted that Microsoft's policies include prohibiting the use of its technology for the mass surveillance of civilians, writing: "We have applied this principle in every country around the world, and we have insisted on it repeatedly for more than two decades," and that the review had found "evidence that supports elements of The Guardian’s reporting. This evidence includes information relating to IMOD consumption of Azure storage capacity in the Netherlands and the use of AI services."

The summary document, published on June 4, 2026, Microsoft states that it took the "reports and allegations seriously" and, following its internal review, it turned to law firm Covington & Burling LLP to conduct an external review.

The document reiterates that Microsoft "does not provide technology to facilitate mass surveillance of civilians," and that due to the company's policy to respect and protect privacy rights of customers, its investigation did not enable it to access IMOD's content and relied on Microsoft business data.

Microsoft stated that its relationship with IMOD was a "standard commercial relationship," but in a footnote added that the company "occasionally provide[s] special access to our technologies beyond the terms of our standard commercial agreements."

"In addition to the commercial relationship with IMOD, Microsoft provided limited emergency support to the Israeli government in the weeks following October 7, 2023, to help rescue hostages." Though details are not provided, Microsoft claimed that this was done on a "considered and careful basis, to help save the lives of hostages while also honoring the privacy and other rights of civilians in Gaza."

The company has put forward five areas in which it will seek to "enhance clarity and effectiveness of our human rights governance and predictability for our customers." These are as follows.

  • Enhancing its pre-contract review process for national security-related engagements
  • Reviewing its existing processes and controls for oversight of security clearances in relevant non-US markets
  • Undertaking periodic reviews of acceptable use and national security-related policies
  • Providing additional guidance to employees regarding Microsoft’s policies related to the acceptable use of its products and services, including how they apply in practice and when and how to escalate related questions
  • Providing additional mechanisms for employees to raise concerns about development or deployment of technology and our commitment.

The company has a new section in its "Microsoft Integrity Portal" which it has called the Trusted Technology Review, and enables employees to raise concerns anonymously.

This is particularly pertinent, as many Microsoft employees over the years have protested against the company's involvement with the Israeli military, and the company has been known to fire employees for staging protests.

The turmoil has led to the formation of a dedicated employee group within Microsoft - No Azure for Apartheid - which has long been calling for Microsoft to cease providing AI and cloud services to Israel.

Nisreen Jaradat, a member of No Azure for Apartheid and former support engineer at Microsoft who was fired following involvement in the worker-led protests, previously spoke to DCD in late March 2026.

Jaradat claims that, prior to her leaving Microsoft, the company had banned the use of words including Palestine, Gaza, genocide, and apartheid in internal emails. "What would happen was if you'd sent an email that included any of those words, or one of our members who disrupted the 50th anniversary celebration, if you used her full name, you would be blocked until somebody manually reviewed it and decided if it was approved or not."

"I was very upset about that. How are you stopping me from saying the name of my homeland, or of what is happening to my homeland and my people?"

Jaradat, in retaliation, sent an email to numerous employees that used the 'censored' words but altered them with spaces or substituted letters to get through the filter. She says she was suspended for six weeks for this.

Microsoft's summary document reiterated that it does not have visibility into how IMOD was exactly using its services, only of its own business data. When asked if she believed this was the case, Jaradat told DCD, "I don't buy it."

Citing The Guardian reporting that suggested CEO Satya Nadella approved the storage of mass intelligence, as well as the fact that it is "known to Microsoft that they are storing massive amounts of audio files, and it is a well-known fact within Palestine that the Israeli military surveils Palestinian phones."

In addition, Palestinian's are required to download an application on their phones (Al Munnaseq) if they want to travel from one restricted area in Palestine to another, which Jaradat says is "under the guise of permit management" and is hosted on Microsoft Azure.

"I think that Microsoft executives themselves had enough information to know that there were massive troves of surveillance data being stored on Microsoft servers."

But beyond that, Jaradat argues: "I think what's even more important is that there is no way to ethically work with a genocidal customer. If I give a pencil to an Israeli militant who uses that pencil to write down the names of who they're going to target, but they don't physically use that pencil to kill anyone. It is still unethical for me to give them that pencil if I know that they are committing genocide and that this will in some way help in their pursuit of committing genocide.

"Microsoft likes to hide behind the excuse of customers owning their data, but as a platform that hosts data, Microsoft has responsibilities, not only ethically, but legally."

Much of the data related to this was reportedly hosted in European data centers, including in the Microsoft region in Ireland.

In December 2025, after Microsoft had ceased some services for Unit 8200, Ireland's ICCL (Irish Council for Civil Liberties) filed a complaint against Microsoft Ireland, which claimed that within days of the original The Guardian investigation being published, Microsoft approved "successive requests from the Israeli Ministry to increase the volume of data it can download from Azure. Then, the Ministry transferred all surveillance data from EU data centers to other infrastructure before an investigation could begin."

Following the publication of Microsoft's "final update," No Azure For Apartheid released a statement: "The notion of Microsoft pretending this is their "final" time being forced to address their complicity is absurd. Nothing short of cutting all ties with the genocidal Israeli military and government will be enough. No Azure for Apartheid will continue to disrupt, protest, and resist until all of our demands are met."

DCD has contacted Microsoft for comment.