something like making progress in hard cryptographic sciences, like finding ways in which hashing function/encryption schemes used today would be rendered easily made useless.
Progress in hardware/cluster computing making things computationally unfeasable made feasable unexpectedly.
Finding backdoors in algorithms being the backbone of the system.
