Microsoft Says There’s ‘No Evidence’ Its Azure and AI Models Have Harmed People in Gaza

Once again, Big Tech is under scrutiny for its role in Israel’s genocide against Palestinians. Recently, Microsoft’s sale of artificial intelligence models and cloud computing services to the Israeli military prompted a series of worker-led protests. But now, Microsoft says there’s no evidence that its products have been used to harm people in Gaza. At least not as far as Microsoft can examine.

On Thursday, Microsoft announced that it conducted internal and external reviews into the Israel Ministry Defense’s use of its products, writing, “We take these concerns seriously.” The company went on to add that it “found no evidence to date that Microsoft’s Azure and AI technologies have been used to target or harm people in the conflict in Gaza.”

Microsoft did not clarify which company it contracted with for external review. Nor did it provide details about the process outside of stating that it included “interviewing dozens of employees and assessing documents.” However, the company added that its reviews are limited. It doesn’t have visibility into how software is used on private servers or on systems outside of its cloud.

Tensions at Microsoft have risen since a February report revealed the extent of its $133 million contract with Israel. According to AP News, Israel’s use of Microsoft and OpenAI technology increased nearly 200 times after Palestinian resistance groups in Gaza launched an attack on Israel on Oct. 7, 2023. The military specifically uses Microsoft’s cloud platform, Azure, to compile information obtained through mass surveillance, like phone calls or texts, which the system transcribes and translates. Overall, it stores over 13.6 petabytes worth of data on Microsoft servers, which, AP News explained, is about 350 times more than what’s needed for the entire Library of Congress.

Last year, Microsoft fired two employees for organizing an “unauthorized” vigil in memory of Palestinians killed in Gaza. In February, Microsoft also kicked five employees out of a town hall meeting for protesting its Israeli contracts. Then last month, Ibtihal Aboussad, a software engineer on Microsoft’s AI Platform team, interrupted the company’s head of AI during a 50th anniversary celebration.

“Shame on you,” Aboussad said. “You are a war profiteer. Stop using AI for genocide. Stop using AI for genocide in our region. You have blood on your hands. All of Microsoft has blood on its hands. How dare you all celebrate when Microsoft is killing children. Shame on you all.”

The Verge reported that Aboussad also sent an email to distribution lists containing hundreds of thousands of Microsoft employees. She wrote, “Microsoft cloud and AI enabled the Israeli military to be more lethal and destructive in Gaza than they otherwise could,” and urged people to sign the No Azure for Apartheid petition, stating, “We will not write code that kills.”

The company’s post comes only a week ahead of a Seattle conference where No Azure for Apartheid intends to protest. In its blog, Microsoft also said that the Israeli military is bound to its conditions of use, which “require customers to implement responsible AI practices” and “prohibits” using its technologies “in any manner that inflicts harm on individuals or organizations or affects individuals in any way that is prohibited by law.”

That assurance falls flat when considering Israel’s track record. Last year, a group of independent human rights experts said that “Israel has openly defied international law time and again, inflicting maximum suffering on civilians in the occupied Palestinian territory and beyond.” That includes murder, torture, sexual violence, forced displacement, bombing vital institutions like hospitals, targeting healthcare workers, journalists, humanitarian workers, and purposefully destroying food systems as a method of war, per a breakdown by the United Nations’ Office for the High Commissioner for Human Rights.

Plus, Israel is committing genocide, which is a war crime. The definition put forth at the Geneva Convention includes specific actions “committed with intent to destroy, in whole or in part, a national, ethnical, racial, or religious group.” Last month, the Gaza Health Ministry reported that deaths in the region now exceed 50,000. In addition, a Reuters analysis found that Israel completely eliminated at least 1,200 families. Although some argue that Israel’s actions don’t meet the condition of “intent to destroy,” its response to the Oct. 7 attacks killed over 5,000 people in the first week, extreme violence that caused a major shift for many experts. South Africa formally brought genocide charges against Israel last year.

Big Tech has provided Israel with support for years, like with Google and Amazon’s Project Nimbus. Microsoft can try to downplay its role by saying that its technology wasn’t directly used for harm. Even if that is true, its technologies make it possible for the Israeli military to expand its destruction of Palestine and its people.