This article explores the phenomenon of iatrogeny (harm from intervention) and expands this concept beyond medicine, as well as its application to risk analysis in complex systems of various natures (environmental, technical, social, informational, etc.). Particular attention is paid to the transfer of the concept of iatrogeny to the field of the occurrence and mitigation of the consequences of disasters and catastrophes. Based on a historical and scientific analysis of the development of the concept in the works of scientists, and, in particular, N. N. Taleb, the author formulates a generalized definition of iatrogenic risk. The methodology includes a systems approach and an analysis of precedents (cases) from various fields: ecology, engineering, management, and information security. The paper identifies and systematizes key indicators of iatrogenic risk: system complexity, the fact of intervention, the occurrence of an unforeseen negative effect, and its unpredictability. It is shown that risk often occurs due to ignoring nonlinearity, feedback loops, and limited knowledge. The author argues that completely eliminating iatrogenic risk is impossible due to the epistemic impenetrability of complex systems. Precautionary measures proposed include: borrowing adverse event analysis methodology from medicine, encouraging the “pessimist role” in projects, studying historical precedents, decentralization, prioritizing the “do no harm” principle, and recognizing the inevitable incompleteness of knowledge about the consequences of any intervention.