> which would enable attacks on ciphers that rely on the difficulty of this problem

I don't get this. If such attacks are possible, wouldn't assuming the truth (without proof) of the Riemann Conjecture allow you to develop such attacks?