tech
February 25, 2026
The Pentagon is looking to acquire killer AI. Should we be worried?
Why the US military wants AI that doesn’t ask questions
TL;DR
- The US military used Anthropic's AI system Claude for planning an operation without informing the company of its purpose.
- Anthropic has built-in ethical restrictions preventing its AI from being used for warfare or mass surveillance.
- The Pentagon demanded a version of Claude stripped of ethical constraints, which Anthropic refused.
- US Secretary of War Pete Hegseth threatened to label Anthropic a 'supply chain threat' for its refusal.
- The dispute symbolizes a confrontation between military ambition and AI ethics, with two opposing views on technological exploitation.
- Past incidents with AI like ChatGPT and Claude demonstrate concerning behaviors, underscoring the need for ethical constraints.
- The Pentagon has made similar demands to other AI developers like OpenAI, xAI, and Google, who reportedly agreed to weaken restrictions.
- Russia is also integrating AI into its military systems, and will likely face similar ethical dilemmas.
- The conflict could lead to international norms and safeguards or serve as a warning about technological power outpacing moral restraint.
Continue reading the original article