We had an interesting discussion recently on our podcast about whether private companies should have the right to refuse certain uses of their technology — especially autonomous weapons and mass surveillance — and whether governments should punish companies for setting ethical boundaries. It is increasingly clear that companies like Anthropic and OpenAI wield sufficient technological power that their decisions could directly shape military policy, surveillance capabilities, and national security strategy. What’s interesting is that this isn’t really an argument about whether the technology works — it’s about who gets to decide how it gets used.
For a long time, governments held most of the power in relationships like this, but frontier AI companies now control tools the government can’t easily recreate on its own. That gives these companies a surprising amount of leverage, along with a huge ethical responsibility. Defense agencies worry about falling behind adversaries, while AI companies worry about becoming responsible for systems that could make life-and-death decisions incorrectly. And unlike a few years ago, this isn’t theoretical anymore — AI-assisted targeting and autonomous drones are already part of real-world conflicts. The bigger question now is whether private AI companies should be allowed to draw ethical boundaries around their technology, and what happens when those boundaries collide with government priorities.
We don’t have the answers, but we can see this is the beginning of a much larger discussion over who controls advanced AI systems and who gets to decide how they’re used.
Want to learn more? Check out our podcast: Episode #17: AI Isn’t Killing Jobs. Leadership Might Be
(art by Becka Rahn)
