Well there are quite a few factors involved in determining the level of interference that will affect a digital signal. Specifically interference occurs at a particular intensity which is usually measured in decibels, the frequency of the interference often dictates what it will and will not interfere with, the amount of background noise not associated with the interference and lastly there are variances in specific materials and configurations being vulnerable to interference.
AC lines can indeed mess with your cat5 cables and you should avoid trying to run them parallel to to your data cables. Generally you won't have much problems with short cables as there is very little attenuation and the signal is bright and crisp. Cat5 cables are optimized to eliminate/reduce crosstalk between the pairs so this makes it far easier to recover a signal at the other end. The longer an AC power line is close to a cat5 cable will increase the amount of interference that is induced in the wire. This induced interference needs to overpower the number of decibels of signal in the cable. Longer cables are more susceptible to interference as the signal is weaker at the end of a long cable run.
Generally speaking interference is introduced into some circuits through passive inductance which is just a fancy name for crosstalk.
Another thing to consider is that a balanced cable is much less likely to experience disruptive interference as both sides of the cable are more or less grounded to bleed off any extra signals via unused wires.
something to consider, you can interfere with any cable if you have the proper intensity, frequency and/or proximity to induce the interference.
EMP is a good example of universal interference or check out a spark gap generator to see what a nasty multi-spectrum noise source looks like.
I could go on and on but for the sake of brevity I will close with a link to more info...
Wikipedia article on Wave Propagation