Somewhere in Google’s Sunnyvale headquarters is a conference room that gained some notoriety in April of last year. Workers entered, took a seat, and wouldn’t leave. They weren’t calling for better pay or benefits. They were calling on their employer to cease doing business with a government that they perceived to be carrying out crimes. Trespassing charges were brought against nine of them. Twenty-eight were let go. Eventually, the figure reached fifty. Even by the increasingly turbulent standards of Silicon Valley, it’s difficult to ignore how unusual that is.
One of the most contentious business agreements in recent memory is Project Nimbus, a $1.2 billion cloud computing contract that the Israeli government, Google, and Amazon signed in 2021. On paper, it appears to be a simple infrastructure agreement that includes digital transformation for government ministries, cloud services, and AI tools. It’s the kind of document that tech companies sign on a regular basis without anyone having to march through the lobby. However, this one is unique, and the causes continue to mount.
Unusual protections were built into the contract from the start. Google and Amazon are contractually prohibited from refusing service to any Israeli government agency, including the armed forces. Boycott pressure prevents them from stopping services. It’s important to take a moment to consider the final clause, which effectively anticipated protest and locked the companies in regardless. It probably depends on who you ask and how kindly you’re feeling to determine whether that was careful legal drafting or something more intentional.
The internal documents that have since come to light are more difficult to ignore. The New York Times reported that at least four months prior to the contract’s finalization, Google’s own attorneys voiced concerns about possible human rights violations. According to reports, they were concerned that Google Cloud services might be connected to Israeli actions in the West Bank. Nevertheless, the business signed. According to a spokesperson, the project covers workloads associated with healthcare, education, and finance; nothing is sensitive or classified. However, internal documents from the Israeli government and Google seem to make that description much more difficult.
Marketing manager Ariel Koren, who openly opposed the project, was given seventeen days to move to São Paulo or risk losing her job. She decided to quit and complained to the National Labor Relations Board and Google’s HR division. Although technically a conclusion, the NLRB’s dismissal of the case due to a lack of evidence may not be entirely satisfying. Many people who witnessed the events felt that the optics were never fully taken into consideration.
The workers who are demonstrating under the banner of No Tech for Apartheid are not a minority. The protest group was joined by more than 200 Google employees. They argue that tools created in American offices may be influencing decisions made thousands of miles away, citing AI-assisted programs purportedly employed by the Israeli military, such as systems that classify buildings, track people, and create targeting lists. Whether Project Nimbus technology is directly linked to those systems is still unknown. However, the contract’s ban on service denial makes it challenging to resolve that ambiguity in a clear and concise manner.
The tech sector has never had to deal with this kind of situation before. For a long time, businesses have sold technology to governments, frequently with little oversight. In a time when engineers know exactly what they’re building and have made the decision, in large numbers, to not, Project Nimbus is testing whether that arrangement can endure. The discussion hasn’t stopped because of the firings. They’ve probably kept it louder than the companies would like, if anything.
