Most countries have fairly stringent 'Data Protection Acts' so if confidential data is being updated on to the backend it's almost impossible to verify the 'compliance' of the data without breaking the various laws governing the storage of confidential data. Unfortunately there is always a fundamental weakness in CLIENT / SERVER systems - it doesn't matter HOW SECURE the server is because an authorised user will always be allowed to update the real data base from some sort of front end terminal. It's almost impossible these days to ban emails in a workplace and confidential emails are often transmitted in scrambled form as people don't want this in plain text all over the Internet. Since ASCII compresses down very much a trick was to convert the file to ascii and compress it before transmission - this made for shorter transmission times - very important back in the days where BAUD rates were still being used - 2400 BAUD (old Hayes Modem) was regarded as a state of the art bit of kit !!!!!. Old XMODEM / YMODEM / ZMODEM protocols for example. Now on the workstation the email message does it's nasty business - it's really a program - and won't be detected by any Virus scanning stuff and there you are -it's only a matter of time before the backend database is corrupted to such an extent it becomes inoperable without a restore etc. You can very easily code a binary file into a bog standard ASCII TEXT message so the email server won't recognize say an illegal inbound.

The usual way is to scan / check email attachments - but that is no good against a LOW TECH attack that doesn't use attachments, macros or anything else in Ms Office or equivalent.

Scrambled ASCII text based stuff doesn't need anything like Word Macros etc etc.

Hitmanpro alert blocks disk format tool