Independent, evidence-first, and deliberately narrow

Human in Control is a public-interest project focused on one question: are increasingly capable AI agents remaining under meaningful human control?

We are not a general AI news site, anti-AI campaign, hype tracker, or broad database of every AI harm. We focus specifically on control failures and the safeguards that prevent or contain them.

We use primary sources wherever possible, clearly distinguish simulations from real-world events, publish uncertainty, accept challenges, and maintain visible corrections.