Story perspectives
Claude: First AI Certified for Military Intelligence in 2025
3/15/2026
1 of 1
Story summary
- In 2025, Anthropic deployed its large language model Claude for national service amid AI safety concerns.
- Anthropic was founded by former OpenAI employees, and Dario Amodei, its CEO, aimed to shape future AI regulations.
- Claude became the first AI certified for classified systems, aiding military intelligence without direct combat capabilities.
- The government contract bars Claude from autonomous weapons or domestic surveillance, and the Pentagon accepted this constraint.
