That 9/11 was in 2001 and the graph (conveniently) starts at 2004?
<benefit of doubt mode>
Graph created in 2014 and supposed to cover ‘last decade’
</benefit of doubt mode>
But even if you did include it and go back to 2001 or earlier it would still be a dwarfed by the gun deaths, and that’s before you even question the idea of whether or not you should include single, infrequent, and unusual events when looking at trends.
Not to mention that they have taken action* to prevent another occurence of a 2001 style event, but aren’t exactly taking much action to address the huge gun toting elephant in the room.
*I’d hate to see the graphs of worldwide/non-US deaths as a result of said action.