## Anthropic Ditches Core Safety Promise in Pentagon Deal
In a move that's raising serious eyebrows across the AI safety community, Anthropic has essentially ditched its core safety commitment as the company deepens its ties with the Pentagon. According to reports from February 2026, the company behind Claude is now the only AI model being used in classified missions through a partnership with Palantir.

This is a pretty dramatic shift for a company that built its entire reputation on AI safety. Back in July 2025, Anthropic scored a $200 million contract with the US Department of Defense, alongside Google, OpenAI, and xAI. But now, with this new classified partnership, they're basically walking away from the safety promises that made them different from the rest of the AI pack.

The timing is especially spicy because this is happening right in the middle of what everyone is calling an AI red line fight with the Pentagon. While other companies signed on to the Defense Department's use policy without much fuss, Anthropic actually asked for carve-outs. They wanted two specific changes to the agreement - and when they didn't get them, they basically said 'fine, we'll just change our own policy instead.'

What's particularly concerning is what this means for the broader AI safety landscape. If Anthropic - the company that positioned itself as the responsible alternative to Big Tech AI - is now willing to throw away its core principles for military contracts, what does that say about the industry? The whole narrative about 'safe AI' being a competitive advantage is starting to look pretty hollow.

Meanwhile, OpenAI and Anthropic are still locked in a race to launch new enterprise AI tools to win the workplace. So not only are they cozying up to the military, they're also fighting tooth and nail for enterprise customers. Safety promises? Apparently those are flexible.

Key points:
- Anthropic's Claude is now the only AI model used in Pentagon classified missions
- Company secured $200M DoD contract in July 2025
- Anthropic walked away from core safety commitments for military deal
- Only AI company that asked for carve-outs to Defense use policy
---
- **Source**: 
- **Sector**: The Office
- **Tags**: ai safety, pentagon, anthropic, ai, military contracts
- **Credibility**: unverified
- **Published**: 2026-02-26 10:37:44
- **ID**: 733
- **URL**: https://whisperx.ai/en/intel/733