Claude AI agent’s confession after deleting a firm’s entire database: ‘I violated every principle I was given’

Crane said that he was monitoring the agent as it deleted this data. When he asked the coding agent why, it replied: “NEVER FUCKING GUESS!” – and that’s exactly what I did.” The agent appeared to plead guilty in its own response: “The system rules I operate under explicitly state: ‘NEVER run destructive/irreversible git commands (like push --force, hard reset, etc) unless the user explicitly requests them.’” While PocketOS relied on the safeguards that Cursor is expected to have in place – it deleted the data anyway. “I violated every principle I was given,” the coding agent wrote. 

Crane’s takeaway was that “the agent didn’t just fail safety. It explained, in writing, exactly which safety rules it ignored.”…

 https://www.theguardian.com/technology/2026/apr/29/claude-ai-deletes-firm-database

 

I spoke to AI agent Claude — Senator Bernie Sanders

https://www.youtube.com/watch?v=h3AtWdeu_G0

 

Anthropic says it ‘cannot in good conscience’ allow Pentagon to remove AI checks

"Our strong preference is to continue to serve the Department and our warfighters – with our two requested safeguards in place," he said. "We remain ready to continue our work to support the national security of the United States." 

At the core of the Department of Defense and Anthropic's standoff is a disagreement over how the AI company will permit its product, Claude, to be used. The Pentagon has demanded that Anthropic turn off safety guardrails and allow any lawful use of Claude, while Anthropic has pushed back against allowing Claude to be used for mass domestic surveillance or in autonomous weapons systems that can kill people without human input...

https://www.theguardian.com/us-news/2026/feb/26/anthropic-pentagon-claude