Artificial intelligence is not neutral code, but always carries values and power structures. The extent of ethical codes worldwide shows diversity and arbitrariness rather than orientation. They often serve as a fig leaf in the form of “ethics washing” or “ethics shopping.” At the same time, the real problems remain: discrimination, surveillance, loss of autonomy, manipulation, or job cuts. An ethics of AI thus easily becomes a checklist that manages and neutralizes conflicts instead of critically reflecting on them. Rights remain empty if they are not embedded in collective practice. The real question is therefore not which principles we tick off, but rather: What kind of world do we want to live in? Responsibility does not arise in committees or papers, but in joint design and politic ...