# Is the random oracle commitment scheme secure against PPT active adversaries?

The probability that a probabilistic polynomial adversary corrupting the sender can finds two pairs $$(m,r)$$ such that the output of the random oracle $$c$$ is the same (break the binding property) is negligible. At the same time, the probability that a probabilistic polynomial adversary corrupting the sender finds a pair $$(m,r)$$ starting from the commitment $$c$$ (break the hiding property) is negligible. Is there any difference between the case in which the adversary is semi-honest or active in this context? The only threat I see is that an active adversary could change the message of the sender, but this is not relevant since the sender could have chosen that message.