You can only exploit a vulnerability, you can't 'turn around' a system without finding some weakness to exploit. Also a lot has to do with how you program (ie. do you check permissions before allowing access to data and objects etc.)
I think dotLRN is as safe the more secure applications and it is easy to work with.
