Tech Employees Rally Against Pentagon's AI Demands in Open Letter
In a significant move, letters purportedly from employees of major technology firms including Amazon, Google, Microsoft, and OpenAI are circulating widely. These communications urge the leadership of these and other tech companies to establish clearer boundaries regarding their collaborations with the U.S. military, specifically addressing concerns over artificial intelligence deployment.
No Tech For Apartheid Leads the Charge
One of the earliest and most prominent letters originates from the group No Tech For Apartheid, an organization known for its past public protests at Google and Microsoft offices. This open letter, issued amidst tensions between Anthropic and the Department of Defense, calls on tech giants to reject Pentagon demands that could compromise AI safety protocols.
The letter states, "We are speaking out today because the Pentagon is demanding that Anthropic abandon two major safety guardrails for Claude, which is the only frontier AI model currently deployed in classified Department of War operations." It characterizes the Pentagon's actions as intimidation, emphasizing the urgency of the situation.
Key Demands and Safety Concerns
On Friday, No Tech For Apartheid released a joint statement titled "Amazon, Google, Microsoft Must Reject the Pentagon’s Demands," representing worker groups from these companies. The statement highlights that their organizations typically advocate for workplace protections, environmental justice, and technology demilitarization, marking this as their first unified declaration.
The Pentagon's ultimatum, as described, pressures AI companies to either comply with its terms or face designation as a supply chain risk, potentially forcing technology provision through the Defense Production Act. The critical safety guardrails at risk include:
- No mass domestic surveillance
- No fully autonomous agents, meaning AI-powered weaponry must retain human oversight
Anthropic has reportedly set a deadline for compliance, but as of Thursday, the company issued a statement rejecting the Pentagon's demands and vowing to uphold these guardrails. The response from the Pentagon remains uncertain, but the letter warns that they may seek alternative models without such protections.
Broader Implications and Corporate Complicity
The letter urges tech companies to refuse compliance if they or their invested frontier labs enter further contracts with the Pentagon. It argues that capitulation could grant War Secretary Pete Hegseth unprecedented powers for mass surveillance and deployment of AI-driven lethal drones, exacerbating existing complicity in atrocities and war crimes.
"It is not a given that our companies will do the right thing," the letter cautions, noting that xAI has already signed a contract with the Pentagon to deploy Grok in classified environments, allegedly without guardrails. Google is in negotiations to deploy Gemini for classified uses, while Amazon and Microsoft have significant investments in Anthropic and OpenAI, with OpenAI also engaging in Department of War talks.
All three companies currently host government data through their cloud services: Google Cloud, Microsoft Azure, and Amazon Web Services (AWS), raising stakes for transparency and ethical oversight.
Calls for Federal Regulation and Worker Action
In the absence of robust federal oversight, the letter advocates for Congress to pass regulations prohibiting irresponsible AI use in violence and surveillance. It concludes with specific demands:
- Executive leadership at Google, Microsoft, and Amazon must reject the Pentagon's advances.
- Provide workers with transparency about contracts with other agencies like DHS, CBP, and ICE.
The statement invites workers to join organizing efforts to prevent the misuse of labor for surveillance, weaponry, and war, underscoring a grassroots push for ethical technology practices in the face of growing military integration.
