900 Google Workers Just Told the Pentagon: Not Our AI

A letter started with 200 signatures on Friday. By Monday, it had 900—nearly 800 from Google and 100 from OpenAI. The title: "We Will Not Be Divided." The target: the Pentagon's decision to blacklist Anthropic as a "supply chain risk."

This isn't just another tech worker petition. It's a direct challenge to how Silicon Valley does business with the military-industrial complex.

The Anthropic Trigger

Anthropic drew a line in the sand. No mass surveillance. No fully autonomous weapons. The company refused to let its AI technology cross these red lines, even if it meant losing government contracts.

The Trump Administration's response was swift: blacklist the company on Friday, then launch airstrikes on Iran hours later. To tech workers, the message was clear—comply or face consequences.

"They're trying to divide each company with fear that the other will give in," the letter reads. Notably, it refers to the Department of Defense as the "Department of War"—a deliberate rhetorical choice that signals how these workers view military AI contracts.

Google's $1.2 Billion Question

Google faces the most scrutiny right now. Reports suggest the company is negotiating with the Pentagon to deploy its Gemini AI model on classified systems—a deal that could mirror Elon Musk's xAI agreement to provide Grok for military use "without any guardrails."

This brings back memories of 2018's Project Maven revolt, when thousands of Google employees protested the company's drone surveillance contract. Google eventually let that contract lapse and created "AI Principles" that explicitly prohibited building weapons.

But those principles have been quietly evolving. Reports suggest Google removed language explicitly prohibiting "building weapons" last year. When pressed for comment on the current Pentagon negotiations, Google has remained silent.

The Solidarity Economy

What's striking about this moment is the cross-company nature of the resistance. Workers at Salesforce, Databricks, IBM, and Cursor have signed letters supporting Anthropic. The "No Tech For Apartheid" coalition is demanding that Amazon, Google, and Microsoft reject Pentagon terms that enable surveillance or other "abusive uses of AI."

This represents something new: tech workers organizing not just within their companies, but across the entire industry. They're essentially saying, "If you pressure one of us, you pressure all of us."

The Guardrails Question

Google's chief scientist Jeff Dean seemed to sympathize with employee concerns, writing on X that "mass surveillance violates the Fourth Amendment and has a chilling effect on freedom of expression." But sympathy from individual executives doesn't necessarily translate to corporate policy.

The fundamental question remains: Can tech companies set meaningful boundaries on how their AI is used, or will market pressures and government demands ultimately override ethical considerations?