The Story: As per a TechCrunch report, Anthropic has pulled back from expanding defense-related agreements with the U.S. government, while OpenAI has proceeded with a formal arrangement.

According to the article, Anthropic’s decision reflects its internal positioning around AI safety and defense boundaries, whereas OpenAI has entered into a structured agreement with U.S. authorities focused on defined, policy-constrained applications.

TechCrunch notes that OpenAI has emphasized safeguards and usage restrictions within its agreement, framing the collaboration as bounded by its existing safety policies. The move comes amid growing momentum within Washington to integrate advanced AI systems into government and national security workflows.

The divergence between the two leading AI labs highlights differing strategic approaches to defense partnerships — one opting for caution and distance, the other choosing structured engagement under defined guardrails.

Why It Matters: As frontier AI systems grow more capable, defense collaboration is becoming a defining strategic decision for leading labs. The split between Anthropic and OpenAI signals that alignment with government institutions may shape not just policy debates, but competitive positioning in the next phase of the AI race.