How AI answers work
What does knowledge cutoff mean?
Knowledge cutoff means the stated boundary beyond which a model's training-based information may not reliably include later developments.
A cutoff is not a promise that all earlier facts are known or correct. Training coverage can be incomplete and uneven.
Search or supplied documents may provide newer information. That does not mean the underlying model has been retrained.
In Seenford
For Seenford, this means judging the answer actually returned, with a receipt linking each verdict to its exact wording.
Example
A model knows an older service name but needs current sources for its replacement.
Related terms
Updated
See what AI says about your brand →