The article explores the ongoing conflict between Washington and the AI company Anthropic, presenting a stark contrast between the vision of Silicon Valley pacifists who advocate for 'safe AI' and the government’s perspective that sees AI as a crucial tool for national security. The refusal of Anthropic to remove the restrictions on its AI model, Claude, in light of a potential U.S. strike on Iran infuriated the Trump administration, which viewed this resistance as an act of 'corporate betrayal' and 'sabotage' at a critical moment. The author suggests that in the context of the U.S. engaged in a relentless competition with China, moral considerations seem to be a luxury that the country cannot afford.

Furthermore, the article raises the question of whether private companies like Anthropic have the right to deny governments access to their products, drawing parallels to traditional ethical dilemmas such as the confidentiality of confessions in religion. In this discussion, the notion of a corporate entity operating independently from governmental authority is examined, seeking balance between ethical stances and national interests. The implications of this clash are profound, as they extend beyond the realm of technology into fundamental debates about corporate responsibility and the role of ethics in the advancement of artificial intelligence.