I don't know how useful comparisons of shark attack vs (insert mundane everyday activity) statistics really are. They are typically not done in a way that is an apples to apples comparison.
As an example, comparing car accident probability to shark attack probability is very much apples and oranges even if the attempt is made to roughly determine number of people going to beaches, ocean, etc. Exposure time would still be extremely different and not accounted for in such an example.
However, divers make up a small portion of shark attacks, although I'm not sure if that is based on normalized data since there are more surface activities and participates than divers.
The best I guess I can provide regarding helping out your friend is the following:
I recently purchased a life ins policy, there were questions about hobbies, sky diving and scuba diving were specifically asked. Then there are specific questions about the type of scuba diving you participate in. After it was determined I was open water recreational only, no overhead environment, no deco diving the premium did not change one bit from the best rate. If insurance companies with tons of data and actuaries ensuring profit are confident I'm not likely to be eaten; why shouldn't I be