A more complete picture here is needed as the discuscion is incomplete(I work in aerospace, not as an engineer, including work for Boeing) Dielectrics are electrical insulators, not conductors. Oddly enough, anodized aluminum is an oxide layer, and aluminum oxide is actually a very good electrical insulator.
Copper itself will only cause galvanic corrosion if it's strictly in an aluminum-aluminum interface. By introducing copper, you turn aluminum into an anode.
If you're attaching steel to aluminum, you've already created a galvanic cell with dissimilar metals. The same physics are at play, absent even copper. This can be seen in daily life with zinc coated nails or flashing. Zinc is consumed as the anode before steel is attacked. Anyone who has worked on old cars or motorcycles can attest to steel bolts seizing in aluminum castings because of this (real, but slow) effect. There's a bunch of powdery aluminum oxide that has corroded in the threads.
In a steel-aluminum interface with copper anti-seize, copper is more noble than aluminum but less noble than steel. Copper doesn't actually facilitate or advance the rate of corrosion. The addition of copper, which lies between the two in nobility, is irrelevant as aluminum will be consumed first, regardless. I mentioned earlier that aluminum oxide (the coating grown by agonizing) is an insulator electrically, so if you scratch through an anodized surface (about 2-3 thou deep) corrossion will happen faster. If you don't, the mechanical interface is fairly sound.
In an aluminum-aluminum interface, zinc anti-seize is used since aluminum is more noble, so zinc is consumed. In a steel-aluminum interface, you want something less noble than aluminum to preserve the aluminum. A product like Jetlube zinc dust petrolatum which conforms to MIL-CID-A-A-59313 is suitable for aluminum-aluminum interfaces or any interface where you want the paste to be sacrificial.
http://en.m.wikipedia.org/wiki/Galvanic_series