I'm assuming that you mean "corrupt data" because of a bad disk (you mentioned bad sectors), and not corrupted data due to a faulty application.
Sometimes yes, sometimes no, depending on the implementation. There are two ways to implement RAID: software and hardware.
The "better" and more expensive way is usually through an IDE/SATA/SCSI controller that has a microprocessor on it that will handle all the functionality invisibly. This means it will do checksums on written data, and allocate the disks in the "best" way for you and so on. If you're using RAID on production/critical data, use this... You probably already know all that though.
Anyway, HW RAID is typically more reliable than OS-level software RAID, and in this scenario that you mention, I don't believe that it will duplicate corrupt data, EXCEPT in a scenario where one of the disks dies, and you rebuild your array using the corrupt disk. Then you'll get bad data being mirrored to your new drive.
Typically though, you won't encounter this scenario because most disks have built-in checksums, the RAID card itself will validate written data with checksums/hashes, as well as OS-level verifications of written data.
Software RAID takes data that would have been written to one disk and simply mirrors the same write functions to a second disk. The same reasoning applies as above as to why bad data might appear, with one added scenario:
The RAID implementation writes to a corrupted disk FIRST, then reads the data back from the corrupted disk, and writes corrupted data onto the second disk.
(make sense? Write to disk0, then read from disk0, then write to disk1)
This would be a piss-poor RAID implementation, IMO, but it may exist since it may speed up the initial write operations, so rather than way for two disks to write, the hook into the disk writing functions from the application is freed faster by only writing to the one, and the RAID software is called in the background to mirror the changes. See what I mean?
Either way, I don't believe that the scenario you're describing is all that likely, considering all of the checks that are in place.