How to “Security Mark” webpages/printouts based on list/library content

Question

I am currently in an environment where we are getting ready to fully deploy SharePoint (WSS 3) to several departments and divisions that have some documents that contain sensitive data. They also may post items to lists/wikis/discussion boards with this same type of data. The problem is that there is a requirement to "mark" this data in such a way as to identify it as this special type and ensure that anytime it is displayed or printed that it be marked accordingly. I know how to develop content types and I was thinking that adding a security content type that maybe just had a yes/no or some other option might work. The main issue is to ensure that the the item is correctly marked when displayed or printed. This also includes list views or exporting. It seems that this is a tough nut to crack but it also seems like we can not be the first to need something like this. Any thoughts/suggestions are welcome!

Edit: I just remembered also that this will require that Titles of entries/uploads will have to contain a security key word which I believe I could do with an event receiver.

Answer 1

In the past I've use a simple content editor web part to designate sub-sites that contain sensitive data. The content type might be a good idea, but it would still be pretty difficult to handle this at the item level without cusotmization because there is no gaurentee that the field will be on every view.

Answer 2

This may or may not solve your problem, but you could do a simple jQuery search on the page for your keywords, and render the appropriate message in the header, title bar, etc. based on if one of those keywords is found or not. Depending on how much content is there it could be a pretty inefficient way to do it, but it would certainly work.

I would have more governance issues with mixing data (like it sounds like you are). In my experience it's a better practice to tag an entire collection as containing/potentially containing export controlled information or some other sensitive type. This way you can actually control access based on permissions that don't cross-contaminate one another.

So while what you're looking for is certainly possible, I'd argue that there's better ways to structure your data thus negating the need to figure out later what each record's security posture is.