In a rare display of cross-company solidarity, hundreds of employees from Google and OpenAI have launched a joint petition urging their leadership to reject the Pentagon’s push for unrestricted access to artificial intelligence. The movement, titled “We Will Not Be Divided,” was sparked by reports that the Department of Defense (DoD) is pressuring tech leaders to strip away ethical safeguards on AI deployment.
The friction centers on Anthropic, a rival AI firm that recently made headlines for refusing to waive its “red lines” regarding military use. According to the petition, the Pentagon is now attempting to play Google and OpenAI against each other, hoping one will yield where Anthropic stood firm.
The “Red Lines” of AI Warfare
The controversy stems from specific restrictions Anthropic has built into its government contracts. These safeguards are designed to prevent two primary scenarios:
- Mass Domestic Surveillance: Using AI to track and monitor citizens within the U.S.
- Lethal Autonomous Weapons: Deploying “killer robots” or systems that can make the decision to take a human life without direct human oversight.
Key Figures and Signatories
As of Friday, the internal movement has gained significant momentum:
- Google: 266 current employees have signed.
- OpenAI: 65 current employees have joined the cause.
- Internal Pressure: At Google’s AI division, DeepMind, over 100 staffers sent a separate, direct plea to Chief Scientist Jeff Dean, asking him to “do everything in your power” to block any deal that violates these ethical boundaries.
Anthropic’s Stance: A Matter of Conscience
Anthropic CEO Dario Amodei has been vocal about the company’s refusal to grant the DoD “any lawful use” of its models. Amodei argues that current AI is simply not reliable enough to manage the life-and-death stakes of autonomous combat.
“AI-driven mass surveillance presents serious, novel risks to our fundamental liberties. To the extent that such surveillance is currently legal, this is only because the law has not yet caught up.” — Dario Amodei, Anthropic CEO
Amodei further emphasized that providing unreliable tools for autonomous warfare doesn’t just threaten civilians—it puts American warfighters at risk.
What’s Next?
The petition serves as a public signal to the Pentagon that the talent pool powering these AI breakthroughs is unwilling to compromise on safety. By standing together, the employees hope to prevent a “race to the bottom” where ethical standards are sacrificed for lucrative government contracts.
