Google, OpenAI workers urge limits on military AI use

More than 200 Google and OpenAI employees have signed an open letter supporting Anthropic’s stance against using advanced AI for domestic surveillance or fully autonomous warfare.

The letter, with verified signatures and some anonymous signers, had over 160 from Google and more than 40 from OpenAI as of 5:30 p.m. on February 26.

The organizers say they are unaffiliated with any AI company or political group.

The letter calls on Google and OpenAI leadership to “put aside their differences and stand together” against Pentagon negotiations seeking agreement to what Anthropic has refused.

Anthropic has said its technology should not be used to surveil US citizens or for autonomous weapons.

🔗 Source: Axios

🧠 Food for thought

Implications, context, and why it matters.

Anthropic remains the only big AI lab resisting the Pentagon’s terms

The Pentagon set a Friday deadline for Anthropic to allow “all lawful use” of its technology, which other labs already accepted for the military’s unclassified systems 1.
Google, OpenAI, and xAI previously agreed to loosen their usual safeguards for use in unclassified military systems 2.
Anthropic faces threats to end its up-to-$200 million contract, brand it a “supply chain risk” (a label Anthropic says is reserved for U.S. adversaries), and use the Defense Production Act, a Cold War-era law, to force compliance 3 4.
More than 200 Google and OpenAI employees signed an open letter backing Anthropic’s opposition to advanced AI use for domestic surveillance or fully autonomous warfare, even though their employers already accepted the Pentagon’s standard for unclassified systems 2.

The fight could split AI companies into defense-aligned and safety-focused groups

The Pentagon wants AI “without ideological constraints,” which is widening the industry divide 5.
xAI and OpenAI agreed to the Department of Defense “all lawful purposes” standard for unclassified systems, plus xAI agreed to deploy Grok in classified systems with no conditions attached, according to the coalition letter 2.
Anthropic says it gave up several hundred million dollars in revenue by cutting off Claude use by firms tied to the Chinese Communist Party or designated as Chinese Military Companies 3.
The employee letter suggests the split runs through individual workplaces, which could strain recruiting across labs.
If the Pentagon labels Anthropic a “supply chain risk,” U.S. defense contractors may need to certify they have no links to Anthropic, which would set a precedent for the tech sector 2.