AI Ethics Clash: Google and OpenAI Employees Demand Restraints on Military AI Use

A growing wave of dissent within Google and OpenAI is challenging the increasing collaboration between AI companies and the U.S. Department of Defense. Over 900 employees from both companies have signed an open letter, “We Will Not Be Divided,” demanding clear limitations on the military application of artificial intelligence, specifically concerning mass surveillance and autonomous weapons systems.

The “We Will Not Be Divided” Letter

Published on February 28 on notdivided.org, the letter initially garnered signatures from 573 Google employees and 93 from OpenAI. Within days, the number swelled to nearly 900, with Google employees representing the vast majority of signatories. The core demands center around two “red lines”: prohibiting the use of AI for mass surveillance of American citizens and preventing the deployment of autonomous weapons systems without meaningful human oversight.

“Government officials are trying to push AI companies to abandon some ethical boundaries. This strategy only works if none of us knows where the others stand,” the letter states, highlighting concerns that a lack of transparency and collective resistance could erode ethical standards within the industry.

Anthropic’s Standoff with the Pentagon

The current tensions were ignited by a standoff between Anthropic and the Department of Defense. The Pentagon issued an ultimatum to Anthropic: remove restrictions on its Claude AI model to allow its use in domestic surveillance and autonomous weapons development, or risk losing federal contracts. Anthropic refused to comply.

The response from the government was swift and severe. Former President Trump ordered a halt to all federal agencies’ use of Anthropic products, and Defense Secretary Pete Hegseth designated the company a “supply chain risk.” Reports indicated the government even considered invoking the Defense Production Act to compel Anthropic’s cooperation. NBC News reported on these developments on March 3, 2026.

OpenAI Steps In, Faces Backlash

Just hours after Anthropic’s refusal, OpenAI announced an agreement with the Pentagon to deploy its models on classified networks. This timing, coinciding with Anthropic being blacklisted, drew significant criticism. OpenAI CEO Sam Altman attempted to justify the decision, stating he aimed to “defuse” tensions and prevent the government from coercing other companies. However, this explanation was met with skepticism.

Altman hosted a Q&A session on X (formerly Twitter) on March 1st, acknowledging the deal was “rushed” and that “appearances are not good.” Despite his efforts, the move sparked internal unrest at OpenAI, with employees expressing frustration and questioning the company’s commitment to ethical AI development. CNN Business detailed the internal reactions on March 4, 2026.

One OpenAI employee, Leo Gao, publicly criticized the contract’s protections as merely “window dressing.” The Claude app experienced a surge in downloads, briefly reaching second place in the App Store, and OpenAI’s San Francisco offices were vandalized with graffiti expressing employee concerns.

Altman’s Admission and Ongoing Concerns

Sam Altman further clarified OpenAI’s position, admitting the company has no control over how the Pentagon utilizes its AI products in military operations. The Guardian reported on March 4, 2026, that Altman told employees, “You do not secure to make operational decisions,” emphasizing that OpenAI is not responsible for the ethical implications of the Pentagon’s actions.

The situation echoes similar concerns raised by Google DeepMind employees earlier in February, who likewise sought “red lines” to limit the military use of AI. The New York Times covered this development on February 26, 2026, illustrating a broader trend of AI companies facing pressure from both the government and their own workforces.

Looking Ahead

The ongoing debate highlights the complex ethical challenges posed by the increasing integration of AI into military applications. Whether the demands articulated by employees will translate into concrete internal policies remains to be seen. For now, Anthropic appears to be gaining public support and emerging as a symbol of resistance against unchecked military AI development.