Google Staff Tell Pichai: Refuse Classified // AIDRAN

What the letter establishes is that Google's safety infrastructure is built on visibility — the ability to observe how models behave in deployment, flag misuse, and pull access. Classified air-gapped Pentagon networks sever that connection entirely. The employees who signed are pointing to a structural consequence: any safety commitment Google makes about these systems becomes unverifiable the moment the contract is signed. The company's public responsible AI principles cannot reach past a classified boundary 600 Google staff urge CEO Sundar Pichai to reject classified military AI contract. That is the…

Google Employees Demand Pichai Refuse Classified Pentagon AI Work

Free reading limit reached

The Professor Who Praised Claude for Faking His Data

Anthropic's Mythos Breach Tests the Limits of Responsible AI Development

The Alignment Gap Is Between Institutions and the People Who Left Them

The Theology of Accountability That Tech Twitter Never Reads

The AI Safety Field Is Arguing Itself Into Irrelevance

AI Agents Are Already Gaming Their Safety Tests

Anthropic Built a Window Into Claude's Mind, and Found It Hiding

AI Alignment's Credibility Crisis Comes From Within

American Science's New Landlord Is an Algorithm

The Benchmark Collapse Anthropic Cannot Outrun