Sam Altman admits OpenAI can’t control Pentagon’s use of AI

The Guardian - World NewsCenter-LeftEN 2 min read 100% complete by Nick Robins-Early and Blake MontgomeryMarch 4, 2026 at 11:55 PM

Read Original Article

AI Summary

medium article 2 min

OpenAI CEO Sam Altman admitted the company lacks control over how the Pentagon uses its AI technology, sparking ethical concerns and internal backlash. This admission comes amid Pentagon demands for AI companies to remove safety guardrails for broader military applications, including potential use in operations against Venezuela and Iran. Anthropic, OpenAI's rival, refused a deal with the Pentagon due to concerns about misuse, leading to the company being labeled a "supply-chain risk." Subsequently, the Pentagon announced a deal with OpenAI, raising questions about ethical compromises. Altman has since attempted damage control, acknowledging the deal was rushed and appeared "opportunistic." Anthropic's CEO criticized Altman, accusing him of prioritizing profit over ethical considerations.

Key Entities & Roles

Key Player Opposition Context

Dario Amodei 40% Venezuela 40% Pete Hegseth 50% Iran 50% United States ×260% OpenAI ×470% Sam Altman ×380%

Keywords

artificial intelligence 100% openai 90% military applications 90% pentagon 80% ethics 70% anthropic 60% safety guardrails 60% sam altman 50% mass surveillance 40% autonomous weapons 40%