seq -w 0 999999999 | sed -E 's/^([0-9]{3})([0-9]{2})([0-9]{4})$/\1 \2 \3/' | awk '{ area=$1+0; group=$2; serial=$3 if (area==0 || area==666 || area>=900) next if (group=="00" || serial=="0000") next printf "%03d-%02s-%04s\n", area, group, serial }'

> In addition to SSNs, the database reportedly includes Americans’ place and date of birth, work permit status, and parents’ names

This is quite a bit more information than just a number.

In the actual report main text, it says that the risk is between 35% and 65%, but does not explain the calculation, if any, that results in those numbers.

It's not until one reaches Appendix A that one finds that this really means that it has been assigned a value of 3 on a scale of 1 to 5, meaning "medium risk", and the value 3 is arbitrarily assigned that percentage range, originating with the U.S.A. FDA's Office of Information Security, where "low risk" (2) is similarly 10% to 35% and "very low risk" (1) is less than 10%.