A session on AI security at Microsoft Build was unexpectedly disrupted this week—not only by protesters, but also by an accidental on-screen leak revealing confidential Microsoft Teams messages about Walmart AI plans with Microsoft. The session, co-hosted by Microsoft’s head of security for AI, Neta Haiby, and head of responsible AI, Sarah Bird, was temporarily derailed when two former employees interrupted the livestream to protest Microsoft’s cloud work with the Israeli government.
During the confusion, the livestream was muted and the camera turned away. But once the disruption ended, Haiby inadvertently switched her screen to Microsoft Teams, briefly exposing internal messages about Walmart’s enterprise AI deployments.
The incident unfolded during a Build presentation focused on best practices in AI security. Protesters Hossam Nasr and Vaniya Agrawal, both ex-Microsoft employees, interrupted the talk with chants accusing Microsoft of complicity in the Israeli government’s actions in Gaza.
“Sarah, you are whitewashing the crimes of Microsoft in Palestine. How dare you talk about responsible AI when Microsoft is fueling the genocide in Palestine,” shouted Nasr, a former employee who was previously fired after holding a vigil for Palestinians outside Microsoft’s headquarters. Both he and Agrawal are affiliated with the activist group No Azure for Apartheid, which has staged multiple protests at Microsoft events.
This marks the third protest interruption at Build this year. Earlier this week, other protests disrupted sessions led by Microsoft’s head of CoreAI and even CEO Satya Nadella during his opening keynote.
Walmart’s AI Plans Exposed on Livestream
After the protest, as Haiby resumed the talk, she inadvertently displayed a live Microsoft Teams chat containing confidential details about Walmart’s deployment of Microsoft’s Entra Web and AI Gateway services.
“Walmart is ready to rock and roll with Entra Web and AI Gateway,” read one message from a Microsoft cloud solution architect. Another quote, attributed to a Walmart AI engineer, stated: “Microsoft is WAY ahead of Google with AI security. We are excited to go down this path with you.”
Walmart is one of Microsoft’s largest enterprise customers, already using Azure OpenAI for various internal AI tools. The leaked chat suggests further integration of Microsoft’s newer AI security frameworks—an area where Microsoft is attempting to outpace rivals like Google Cloud and AWS.
Microsoft declined to comment on either the Build protest or the leaked Teams messages when contacted.
Microsoft’s Relationship With Israel Under Scrutiny
The protests come just days after Microsoft released a statement asserting that an external review found no evidence that its AI or cloud services have been misused in the war in Gaza. The company claims its partnership with Israel’s Ministry of Defense is “a standard commercial relationship” and says it remains compliant with its AI Code of Conduct.
Still, protest groups and former employees continue to raise ethical concerns about the opacity and accountability of Microsoft’s government contracts, especially in conflict zones.
Agrawal, one of the protestors, also interrupted Microsoft’s 50th anniversary event last month during a panel featuring Bill Gates, Steve Ballmer, and Satya Nadella. She was let go shortly after submitting her resignation.
Between high-profile leaks and high-tension protests, Microsoft Build 2024 has faced an unusual level of disruption. With increased scrutiny over both its ethical AI commitments and cloud contracts with governments, the company is navigating a delicate balance between innovation, transparency, and trust.
The inadvertent exposure of Walmart’s AI integration plans—amid a protest centered on AI ethics—highlights just how complex that balance has become.
