The Pentagon uses extortion to get an American company to change their existing contract. This podcast details a 2026 conflict between the Pentagon and the AI company Anthropic regarding the military's use of the Claude model. According to the report, the government designated the firm a national security threat after its CEO refused to remove safety restrictions against domestic surveillance and autonomous weaponry. While the Department of War claimed this was a necessary step for operational reliability, the author argues it was a coordinated campaign to replace an uncooperative partner with more compliant competitors like OpenAI and xAI. The narrative highlights a significant contradiction, noting that the military utilized Anthropic’s technology for airstrikes in Iran immediately after labeling the company a risk. Ultimately, the source characterizes this event as the end of Silicon Valley's independence, illustrating how the state can use economic coercion to force private tech entities into unconditional submission.