Google, OpenAI Staff Back Anthropic's DoD Stance in Letter - AI News Today Recency

# Google, OpenAI Staff Back Anthropic's DoD Stance in Open Letter Amid Escalating AI Ethics Clash

In a bold show of industry solidarity, over 200 employees from Google and OpenAI have signed an open letter supporting Anthropic's refusal to grant the U.S. Department of Defense (DoD) unrestricted access to its AI models, particularly for mass surveillance and autonomous weapons.[1][2][3] As a Pentagon deadline looms, the dispute has ignited a fierce debate over AI's role in military applications, with Anthropic CEO Dario Amodei standing firm against what he calls unreliable and undemocratic uses of frontier AI.[1][5]

Pentagon's Ultimatum Sparks AI Industry Rebellion

The conflict erupted as the Pentagon issued a "best and final" offer to Anthropic, demanding removal of restrictions that bar its AI from powering fully autonomous weapons or large-scale surveillance of Americans.[1][2] Defense Secretary Pete Hegseth warned of invoking the Defense Production Act (DPA) or labeling Anthropic a "supply chain risk" if it refuses, with a Friday deadline intensifying tensions.[2][4] Anthropic rejected the terms, arguing its models like Claude are "not reliable enough" for autonomous lethal decisions and could erode democratic values by stitching public data into invasive citizen profiles.[1][5]

Pentagon officials escalated rhetoric, branding Amodei a "liar with a God complex," while Anthropic highlighted its pioneering deployments in classified networks for intelligence and cyber operations—excluding only the two red lines.[1][5] Rivals like OpenAI and xAI have agreed to "all lawful purposes" clauses, but employee pushback signals growing unease.[1][4]

Employee Petition Gains Momentum Across Tech Giants

More than 220 verified signatures from Google (176) and OpenAI (47) back Anthropic, urging leaders like Google's Jeff Dean to adopt similar limits on military AI use.[1][2][3] The letter accuses the DoD of divide-and-conquer tactics, negotiating separately with Google and OpenAI to secure what Anthropic denied, and calls for unity against "domestic mass surveillance and autonomously killing people without human oversight."[2][3][4]

Organizers emphasize independence from AI firms or politics, with options for anonymity.[3] OpenAI CEO Sam Altman echoed support in a staff memo and public comments, opposing DPA threats and sharing Anthropic's red lines, while a spokesperson confirmed alignment.[1][4] Google employees separately petitioned their chief scientist for Gemini model safeguards.[1]

Anthropic's Ethical Stance and Broader Implications

Anthropic has proactively supported U.S. national security, being first to deploy models in classified settings and national labs for non-controversial tasks like planning and simulations.[5] CEO Amodei stressed the DoD's prerogative to choose contractors but vowed a smooth transition if offboarded, prioritizing safeguards over unrestricted access.[4][5]

The standoff threatens industry-wide rebellion, with over 330 from Google DeepMind and OpenAI also endorsing Anthropic.[1] Congressional allies and groups like Common Cause demand oversight, urging testimony on "all lawful purposes" and AI risks.[6] This tests whether employee activism can reshape corporate-military deals, especially after Google's 2025 policy shift.[3]

Reactions from Leaders and Watchdogs

Company executives show sympathy: Altman criticized Pentagon pressure, and informal signals suggest alignment.[4] Meanwhile, watchdogs like Common Cause rally Congress to probe DoD overreach in a $200 million contract fight.[6] The petition warns of solidarity's power: "This letter serves to create shared understanding... in the face of this pressure."[2]

Frequently Asked Questions

What are Anthropic's main red lines with the DoD? Anthropic refuses to allow its AI for mass domestic surveillance or fully autonomous weapons without human oversight, citing reliability issues and democratic risks.[1][5]

How many employees signed the supportive open letter? Over 220 from Google and OpenAI, with figures like 176 from Google and 47 from OpenAI; broader counts exceed 330 including DeepMind.[1][2][3][4]

What penalties does the Pentagon threaten against Anthropic? Invocation of the Defense Production Act to force compliance or designation as a "supply chain risk," potentially barring business with DoD-linked firms.[2][4]

Has OpenAI or Google officially responded to the letter? No formal responses yet, but OpenAI's Sam Altman and a spokesperson voiced support for Anthropic's limits; Google has not commented publicly.[1][4]

Why does Anthropic still work with the military? It deploys models for intelligence, cyber ops, and planning in classified settings but draws firm lines on surveillance and lethal autonomy.[5]

Could this affect other AI companies' DoD contracts? Yes, the letter warns of DoD negotiations pressuring Google and OpenAI to drop similar safeguards, potentially sparking wider resistance.[2][3]