Microsoft Search, Ask Us Anything

Question is generic. As enterprise we have multitude of systems in which valuable information is contained. We plan to expose that to our end-users in transparent manner without require them in which system the information resides. As means via Microsoft Search, and the content be included in index via connectors: standard if available, build custom if not. But we need to ensure adherence to access permissions on the original information being crawled. What is required on a source system to have it crawlablem and on query time respect in the returned result set the permissions on original data.