AI Workers Unite Against Pentagon Demands for Autonomous Warfare and Surveillance

Employees from Google and OpenAI have launched an open letter in solidarity with Anthropic to resist Department of War pressures. The Pentagon is allegedly threatening Anthropic with legal retaliation for refusing to allow its AI to be used for autonomous killing and domestic surveillance. The signatories urge their respective leadership to maintain ethical 'red lines' and reject military demands that compromise human oversight and privacy.

Key Points

• The Department of War is threatening Anthropic with the Defense Production Act for refusing to tailor its models for military surveillance and autonomous killing.
• The Pentagon is attempting to divide the AI industry by negotiating with Google and OpenAI to accept terms that Anthropic has rejected.
• Over 600 employees from Google and OpenAI have signed a letter of solidarity to prevent their companies from being played against each other.
• The signatories demand that their leaders maintain ethical boundaries against domestic mass surveillance and lethal AI use without human oversight.
• The letter serves as a tool for shared understanding and collective resistance against government pressure to weaponize AI technology.

Sentiment

The community is overwhelmingly sympathetic to Anthropic and the AI workers' position, strongly critical of the government's retaliatory approach. Most commenters support the right of AI companies to maintain ethical red lines and view the supply chain risk designation as an authoritarian overreach. However, significant cynicism exists about whether resistance will ultimately change anything, and a vocal minority argues the government has legitimate legal authority to compel cooperation.

In Agreement

• The 'supply chain risk' designation is a devastating economic weapon that forces all government contractors to sever ties with Anthropic, going far beyond simply not contracting with them
• AI models are not reliable enough for autonomous warfare and could endanger both warfighters and civilians
• Mass domestic surveillance of Americans violates fundamental constitutional rights
• Using procurement rules as political punishment sets a dangerous precedent that could be applied to any US company that displeases the administration
• Strong-arming top AI scientists and engineers will lead to brain drain as talent flees to Europe and the UK
• Historical parallels to fascist regimes aligning industry with state interests suggest this is part of a broader authoritarian pattern
• The right of companies to maintain ethical red lines is analogous to conscientious objection — corporations should not be stripped of moral agency

Opposed

• The Defense Production Act gives the government legal authority to compel cooperation from companies whose technologies are relevant to national security
• Corporations should not have broader rights than individuals to refuse government service, especially when individuals can be drafted into war
• Anthropic previously lobbied for regulations restricting open-source AI models, making their resistance to government control appear hypocritical
• This is 'business as usual' — the US government has always coerced tech companies, as demonstrated by PRISM and Yahoo's forced compliance
• Companies that signed contracts with the Pentagon do not get to unilaterally dictate which applications are acceptable
• The government can simply find a willing provider — the market will produce companies without these ethical reservations