Responsibility to the Problem: An Explicit Standard for the Expert–AI System

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Discussions of artificial intelligence in science ask who is responsible when AI produces an incorrect result. This question is vague unless it identifies the object of responsibility, evidence, decisions, and cognitive process. Public criticism often isolates an erroneous output while omitting the question, information, corrections, verification attempts, and human decision to accept it. This paper reformulates responsibility for the work of the Laboratory of Information Systems and Laboratory of Military Ecotones. The relevant unit is the expert–AI system: problem formulation, evidence, AI system and version, alternatives, expert corrections, verification, and final decision. Cognitive work may be joint, but accountability remains human because the expert decides whether results are accepted, published, or applied. The framework distinguishes five levels: personal cognitive, epistemic, process, decision, and practical responsibility. These concern genuine engagement with the problem; separation of observation, interpretation, hypothesis, and verified fact; reconstruction of the cognitive path; identification of the decision-maker; and contribution to recognizing or resolving war-related environmental contamination. Documentation combines an archived interaction with a Logical Record of Expert–AI Problem Solving. Responsibility is thus evaluated through an explicit, reconstructable, and correctable cognitive process rather than rhetorical fear of AI assistance.