dear authors,

after checking the referees' reports, the discussion section and your submitted manuscript, we believe that your paper has potential interest for the journal readers but it is still in a form which is not acceptable by the journal, due to some issues that are quite important.
For this reason, we will reconsider a new, revised submission should you be ready to incorporate some major changes, as requested by the two referees.
In particular, I feel that some of their requests are particularly reasonable and, therefore, mandatory to bring the manuscript up to the journal standards. I refer to:

1. the need to clarify the methodology with respect to the (many) existing standards existing in landslide hazard studies

2. the need to clarify, homogenize and standardize the taxonomy, the ontology and the terminology used (including the way you refer to the landslide object inside a classification method which is rather different from the ones widely accepted

3. the need to detail some parts of the methodology by adding figures and documentation (e.g. flow chart and expert rules used)

4. the need to try to quantify the impact of the different possible choices of rules and settings on the final result

5. the authors should be more careful in stating quite uncertain and very challenging objectives, such as those of unifying EU INSPIRE standards in AI intelligence applications and terminology since those standards are still debated and non-existing at the moment. Moreover, they cannot be defined by force since they require a general agreement among the many institutions and consortiums working on landslides and related risks. The task is overwhelmingly complex and the authors do not seem to clarify what is the contribution of their work towards this general objective

6. the need to better frame the entire method and results within the existing landslide risk mitigation framework (by also citing the most relevant literature). When the proposed tools would be actually usable and useful? Possibly, where no accurate landslide inventory maps are already available? And under which conditions and pre-requisites?


Please also answer to all the remaining minor issues and specific comments highlighted in the reviewer's reports, one by one, by modifying the manuscript where required and directly replying to all remarks and comments on a separate document to attach to the new modified manuscript.

I am confident that, after that, the second review round might be expeditious and more successful.

Thank you for submitting to NHESS,
best regards
FC

dear Authors,

I am happy to announce that, after the last revision round, your paper has been judged as acceptable for publication, provided that you can incorporate a few small technical corrections, as highlighted by the two reviewers (see attached reports).
In particular, you should concentrate on:

1. improving the graphical quality of some of the figures
2. check caption of Fig. 5
3. check the visibility of Table in appendix B1
4. if possible, change the text according to the suggestion of reviewer #2

After that, your manuscript will be directly redirected to editorial production.
I congratulate you for the successful submission and I hope this paper could fruitfully contribute to the large and still uncertain discussion on the usage of standards in AI applications for natural hazards.
Thank you for choosing NHESS,
best regards
FC