Under Secretary Emil Michael just dropped a bomb on effective altruism (EA) in AI governance. His core argument: EA-driven AI companies are trying to position themselves as equals to democratically elected governments in transnational AI regulation frameworks.
The technical implication? This isn't about code or models—it's about who gets to set the rules for AI deployment at scale. EA proponents have been pushing for global governance structures where private AI labs sit at the table alongside nation-states, effectively giving unelected tech CEOs veto power over AI policy.
Michael's team at the Department of War identified this early and claims they're "almost done taking them out" of the department. Translation: They're purging EA-aligned advisors and contractors who were embedding this governance philosophy into military AI strategy.
Why this matters for builders: If you're working on AI systems with government contracts or dual-use applications, the ideological alignment of your company's leadership is now a security clearance issue. The pendulum is swinging hard away from "AI safety through corporate-government partnership" toward "democratic accountability over technocratic control."
The real question: Can you build AGI-scale systems without some form of coordinated international oversight? And if so, who gets to decide the rules—elected officials or the people who control the compute?
The technical implication? This isn't about code or models—it's about who gets to set the rules for AI deployment at scale. EA proponents have been pushing for global governance structures where private AI labs sit at the table alongside nation-states, effectively giving unelected tech CEOs veto power over AI policy.
Michael's team at the Department of War identified this early and claims they're "almost done taking them out" of the department. Translation: They're purging EA-aligned advisors and contractors who were embedding this governance philosophy into military AI strategy.
Why this matters for builders: If you're working on AI systems with government contracts or dual-use applications, the ideological alignment of your company's leadership is now a security clearance issue. The pendulum is swinging hard away from "AI safety through corporate-government partnership" toward "democratic accountability over technocratic control."
The real question: Can you build AGI-scale systems without some form of coordinated international oversight? And if so, who gets to decide the rules—elected officials or the people who control the compute?