Pentagon has reportedly sought clarification from Anthropic over safeguards surrounding its AI chatbot Claude, as concerns grow about how advanced generative AI systems could be used in military contexts.