I\'ve looked at the HISA plan of action following the fatalities at CD, Laurel, and Saratoga as well as the report on the fatalities at CD. It appears that HISA has not looked, and does not intend to look, for connections between move-ups and injuries. And if they don\'t include some form of performance data in their database, even with AI they are not going to find any connections if they exist.
Have you thought of offering some part of your database (leaving out the most recent year or two to prevent anyone from using it to build their own) to a grad student in data analysis looking for a thesis project? I doubt you have fatality data in the database, but looking at horses that DNF\'d in their final start might be a good proxy.