Proximate cause
The last event before the harm, closest to the output (for example, "the model produced a one-sided answer"). Real and worth recording, but where investigation begins, not ends.
Defined in 2 GAGE programs, which carry 3 distinct definitions of it. The wording above is taught in AI Governance: Applied Mastery.
How each discipline defines it
The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.
Also called legal causation. The doctrine that limits how far back a court will trace responsibility even when a party's conduct is a genuine but-for cause of the harm, because courts generally stop assigning liability once an outcome becomes too remote or unforeseeable from that party's original conduct.
The last event before the harm, closest to the output (for example, "the model produced a one-sided answer"). Real and worth recording, but where investigation begins, not ends.
The condition or event that occurred immediately before a failure and directly triggered it. Almost always true as stated, and almost never sufficient by itself as an explanation, because it describes what happened without explaining why the system allowed it to happen.
Where it is taught
The exact lessons this term appears in. The first 7 topics of every program are free with a free account.
- The post-incident review: what you missed, written without blaming the tool · Shipping AI and Surviving the Incident, AI Governance: Applied Mastery
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.