Russian practice: who is held responsible for AI
Russian practice differs from the international record: fewer disputes about models, more about responsibility. We follow the courts, regulators and agencies that rule on what an AI system has done.
Courts fine parties for case law a model invented. Universities have expelled students over work written by a machine. Employers hand duties to software and answer for the outcome, while the law says little about the tool itself.
A pattern runs through the record: courts protect the person when a procedure was broken, and do not ban the technology. Restrictions come from below, from agencies and regions; adoption, on the other hand, comes from above.
Not every entry has a primary source. Where one is missing, the entry is marked as unverified, and the mark is not decoration: such a case cannot be cited in an official document.