A hazard is not a risk.
A risk is not a hazard.
What is a hazard?
A hazard is not a risk. A risk is not a hazard. See the briefing on "risk".
Every professional risk assessment begins with hazard identification. A hazard is the inherent potential of something to cause harm. It is the source of the danger, independent of whether or with what probability harm will actually occur.
For example, a shark in the ocean is a hazard. The hazard exists whether a swimmer is nearby or not; it is a general potential for harm. A "risk" materializes once the swimmer is exposed to the hazard, for example, by swimming near the shark. This action creates a concrete scenario where the probability (P) and severity (S) of harm can be determined.
Despite this definition, "hazard" is frequently confused with "risk," even in professional circles. This confusion is fueled by everyday language, but also by standards, repositories, and regulatory use where the terms can be ambiguous. The EU AI Act itself is a prime example, particularly in its use of terms like "high-risk" and "systemic risk".
Recommendations
The term "risk" is often confused with "hazard". If you have the impression that another person confuses them, actively ask for clarification.
When writing or speaking about risk and/or hazards in relation to AI technologies, define the terms and use them coherently. Don't add to the confusion.
The mere existence of a hazard (e.g., a powerful AI capability) is not a reason to stop an innovation project. Always conduct a thorough risk assessment to determine the actual risk. A low-probability, low-severity risk may be perfectly acceptable.
Don't wait for an incident to reveal a hazard. A core part of professional risk management is to systematically scan for and identify potential sources of harm (i.e., hazards) before they appear.
The glossary from the United Nations Office for Disaster Risk Reduction (UNDRR) provides an excellent and authoritative definition of "hazard," including its different types (e.g., natural, technological, and biological). Consulting expert sources is a core practice of professional risk management.
Author
Claudia Otto
As a lawyer and researcher, Claudia specializes in AI safety, security, and risk assessment under the EU AI Act, the subject of her Master's thesis in Security and Disaster Management (MBA).
Need guidance on risk and compliance?
Cite this briefing
Otto, AI Risk Literacy, What is a hazard?, September 2025