3.7 Defined term
hybrid AI
A method that combines mechanisms from different families: learning approaches, knowledge forms, neural designs, symbolic reasoning or statistical models.
What this settles
The definition acknowledges that many real systems are combinations. It removes the false choice between a single label and no description, and it points the reader to the seams.
Who it binds
No one. It is a definition. It becomes operative when a system is described as hybrid and a claim depends on the combination.
What to do in your practice
Record which families are combined and what passes between them. Name the authoritative component for each decision and test a fallback. A hybrid label without the seam detail says almost nothing.
Evidence an auditor or regulator asks for
| A seam map naming the families and the interfaces. | Gives the hybrid label its content. |
|---|---|
| A fallback test with the observed result. | Shows the combination behaves as claimed. |
| An authority note per decision. | Answers which component decides. |
Which families are combined here, and what happens at the seam when they disagree?
Common gaps this page catches
- Hybrid is used as a synonym for complex, with no families named.
- The seam is undocumented and a failure is attributed to the wrong component.
- No fallback is tested, so the combination is assumed to be robust.
What to prepare before publication
Name the families and draw the seam map. It is usually the fastest way to find the untested path in a modern system.
How to test it here
Take one live system rather than a template, find the field or sentence where hybrid AI is used, and put this page beside it. Record the answer, the artefact and the date. A glossary entry on its own is an intention; a dated record beside a live artefact is evidence.
What this page does not claim
The page does not reproduce the draft, does not know the descriptor wording that the public preview withholds, and does not certify the system or the organisation. It states the unit at the claim level "defined term" and stops where the public record stops.
Common questions
- What does hybrid AI mean in ISO/IEC DIS 42102?
- A method that combines mechanisms from different families: learning approaches, knowledge forms, neural designs, symbolic reasoning or statistical models.
- What evidence should we keep for 3.7?
- A seam map naming the families and the interfaces. A fallback test with the observed result.
- Is hybrid AI a requirement?
- No. The draft is a descriptive framework. It sets no management system requirement and no conformity test. It becomes binding only when a contract, a regulator or a management system adopts it.
Related units
Last updated: · Markdown version · Check your system