NO ONE INTENDED THIS : On How the Big4 Harm Humanity Without Intending To, Why the Absence of Intent Is Not Absolution, and the Six Moves That Break the Loop

Every account of AI harm falls into one of two registers. The first is accusatory: they know,they profit, they don't care. The second is exculpatory: no one could have predicted this,they're trying their best, the technology is new. Neither register survives contact with the evidence. The accusatory register fails because the Big4 employ serious safety researchers, publishreal risk assessments, and have disclosed genuine hazards in SEC filings. The people arenot villains. Most of them are worried. The exculpatory register fails because the harms are documented, dated, and in some cases disclosed by the companies themselves. A harm that is known and continued is not an accident. This paper states a third register: structural harm. The harms arise from the architecture, the incentives, and the coordination equilibrium — not from intent. But the absence of intent does not dissolve the harm, and it does not dissolve the responsibility, because the parties know what they are producing and have not named a condition under which they would stop. And because the harm is structural, the way out is structural too. The paper closes with six moves that break the loop — each available to a diferent actor, each able to begin this week, none requiring permission from the Big4. Five of them make the alternative visible.The sixth makes the deferred cost arrive on time.