Every risk family in the pack

One family (tariffs) was read by the model in this run. The other 13 are keyword counts only: how many filers use the words, not how many name the risk. Every number is on the same denominator of 4,551 comparable filers.

"Mention" is a keyword match and costs nothing. "Names it" and "first time" come from the model reading both years, for the families this run read.
FamilyMentionNames itFirst timeDid not
Tariff exposure3,8413,731333820
Regulatory4,534not read in this runnot read in this runnot read in this run
Financing4,446not read in this runnot read in this runnot read in this run
Litigation4,414not read in this runnot read in this runnot read in this run
Cyber incident4,407not read in this runnot read in this runnot read in this run
Supply chain4,392not read in this runnot read in this runnot read in this run
Geopolitical4,218not read in this runnot read in this runnot read in this run
Climate4,166not read in this runnot read in this runnot read in this run
Labor4,042not read in this runnot read in this runnot read in this run
Key person3,934not read in this runnot read in this runnot read in this run
Going concern3,705not read in this runnot read in this runnot read in this run
Intellectual property3,598not read in this runnot read in this runnot read in this run
AI dependency3,329not read in this runnot read in this runnot read in this run
Customer concentration1,284not read in this runnot read in this runnot read in this run

Mentioning a risk in a list of things that could go wrong is not naming it as a risk to the business, which is why a family is only counted as named once the model has read it. How the distinction is drawn. To have another family read in the next edition, ask for it.