The Pentagon Wants AI Companies to Help Build Weapons. They Said No. Now It's Getting Ugly.
The U.S. military is pressuring Anthropic, OpenAI, and Google to remove safety guardrails so AI can be used for weapons and battlefield ops. The companies are pushing back.
This is a big one. The Pentagon wants to use AI for weapons development, intelligence gathering, and battlefield operations. Makes sense from a military perspective. The problem? The companies building the best AI don't want their tech used that way.
Anthropic, OpenAI, Google, and xAI all have safety rules built into their AI systems. Things like "don't help build weapons" and "don't assist with surveillance." The Pentagon is essentially saying: remove those rules or we'll take our contracts elsewhere.
This puts AI companies in a tough spot. Military contracts are worth billions. But their entire brand is built on being "responsible" AI companies. You can't market yourself as the safe AI option while helping design missiles.
It's the biggest tension in tech right now: national security vs. AI ethics. And there's no easy answer.
As reported by CNBC.
Sponsored