As governments race to adopt artificial intelligence for national security, not every AI company is willing to give unrestricted access. Anthropic CEO Dario Amodei states that the company is open to collaborating with the U.S. military on most AI applications, including intelligence analysis, logistics, and operational planning. But he has drawn two firm red lines. First: Domestic mass surveillance. Using AI to analyze large-scale data on civilians, particularly data acquired from private companies, raises significant legal and ethical concerns. Amodei argues the technology is advancing faster than the laws meant to govern it. Second: Fully autonomous weapons. Not assisted systems. Not human-guided drones. Weapons that could select and fire on targets without any human involvement. According to Amodei, today’s AI systems are still too unpredictable for that level of responsibility, and there has yet to be a serious public conversation about how such systems would be overseen. The...