Which extractor does NOT have a retraining activity?

The correct answer is based on the nature of the extractors mentioned. The Regex Extractor and Form Extractor do not require a retraining activity. The Regex Extractor relies on predefined regular expressions to identify patterns in the data, which means it operates based on rules established beforehand rather than learning from new data. Therefore, it does not have a retraining mechanism.

Similarly, the Form Extractor's setup involves specifying the structure and patterns directly, allowing it to work with any document that conforms to that structure without the need for retraining.

Forms AI, on the other hand, utilizes machine learning models that can improve their accuracy based on new data, and hence, it has a need for retraining to adapt to changes or improvements in data handling over time.

Thus, the correct selection identifies that both Regex and Form Extractor are capable of functioning without any retraining activities, distinguishing them from the Forms AI extractor which does require updates to its model through retraining.