The Microsoft Build 2025 conference was surprised when a demonstration interrupted a session on AI security, unwittingly revealing Walmart’s plans for its forthcoming major forays into artificial intelligence.
The session, presented by Neta Haiby, Head of AI Security at Microsoft, and Sarah Bird, Head of Responsible AI, was supposed to concentrate on how to protect AI systems. But it was temporarily disrupted when two ex-Microsoft staff, Vaniya Agrawal and Hossam Nasr, ran on stage. The duo, now engaged in the “No Azure for Apartheid” campaign, were objecting to Microsoft’s cloud deals with the Israeli government. Nasr had yelled, “Sarah, you whitewash the crimes of Microsoft in Palestine, how could you speak about responsible AI when Microsoft is driving the genocide in Palestine?” Security promptly got rid of the protesters, but the disruption had surprising repercussions.
During the chaos, Haiby accidentally shared her Microsoft Teams screen with the live Build audience. What flashed on the screen momentarily was an internal chat showing Walmart’s plans to embrace Microsoft’s Entra Web and AI Gateway services. One of the Microsoft cloud solution architects posted, “Walmart is ready to rock and roll with Entra Web and AI Gateway.” Another post, allegedly made by a Walmart AI engineer, said: “Microsoft is WAY ahead of Google with AI security.”. We are excited to go down this path with you.”
The screen share also revealed internal commentary on Walmart’s current AI tool, “MyAssistant,” which runs on Azure’s OpenAI Service. Launched the previous summer, MyAssistant is designed to help Walmart employees generate marketing content and summarize internal documents. But the messages also raised concerns, describing the tool as “overly powerful and needs guardrails, —highlighting ongoing challenges around responsible AI usage. According to a report by The Times of India, the messages “underscored concerns over Walmart’s current ‘MyAssistant’ tool, the messages framed as being ‘too powerful and needs guardrails.’
Walmart has been a prime partner of Microsoft’s AI network for quite some time, employing Azure OpenAI on numerous internal ventures. The recently revealed information reveals a far more extensive collaboration, with Walmart’s engineering staff having a high degree of faith in Microsoft’s technology, particularly when pitted against the AI security solutions offered by Google.
This was not the first time the conference had been disrupted. Agrawal and Nasr’s protest was the third interruption at Build 2025 following previous demonstrations during the CEO keynote address and other sessions. Both have been previously let go by Microsoft for organizing Palestine solidarity activities, and have since emerged as leading voices in an expanding movement calling on tech firms to rethink contracts with governments entangled in contentious geopolitical struggles.
Microsoft has thus far refrained from commenting on both the protest and the inadvertent leak of Walmart’s strategy. The Times of India mentioned, “Microsoft has not commented on the Build disruptions or the accidentally revealed Walmart partnership details.”
Though the protest itself was brief, its shockwaves had a profound impact. It gave us a glimpse into the delicate negotiations driving high-level enterprise AI collaborations—and how public activism can meet private business in a convergence that resonates far beyond the event horizon.
