Quantum entanglement doesn't transmit information.

When you have two experimenters (usually called Alice and Bob, because real people aren't named A and B) doing widely-separated experiments on entangled particles, each of them does a separate experiment on each particle. For example if the entangled particles are photons with opposite polarizations, each experimenter must decide whether to analyze the polarization as horizontal/vertical, or one-or-the-other 45º diagonal, or left/right circularly polarized. What Alice and Bob get from these experiments in their own labs are completely random lists of orientations (horizontal, +45º, horizontal, right-circular, vertical, –45º, etc.).

It's only when Alice and Bob compare notes later — using a classical communication channel, like a telephone — that they can determine whether there was entanglement or not. Alice and Bob must compare notes to determine which measurements they made using the same basis. If Alice determined that her 47th photon was horizontally polarized, and Bob determined that his 47th photon was right-circular polarized, those two measurements are completely uncorrelated and carry zero information. However, if Alice determined that her 48th photon was horizontally polarized, and Bob measured his photon in a horizontal/vertical basis, then Alice knows (assuming a high-fidelity experiment, low noise, etc.) that Bob's 48th photon must have come out vertically polarized.

This is great for encryption! It lets Alice share innocuous information (her detector orientations) over an insecure communication channel, so that Bob suddenly knows secret information (a random subset of Alice's measurement results) about Alice's lab. Even better for encryption, if a third party (Eve, the eavesdropper) tries to "steal" the secret information by measuring Bob's photons en route, the entanglement is destroyed in a detectable way. But Alice can't control her measurement results to send information to Bob. Even if Alice and Bob agree in advance on the pattern of measurements they will make, e.g. "always horizontal/vertical" or "diagonal every third time," the polarization results in Alice's lab are still random. She can know what Bob is getting, but cannot send him a message.

The philosophically interesting part of quantum entanglement is even more subtle. If Alice and Bob measure their polarizations in different but not orthogonal bases — for example, Alice measures horizontal/vertical and Bob measures linear polarization at 22.5º rather than at 45º — then quantum mechanics predicts more correlation than any classical theory. But from your perspective I think just means that even more measurements are required to determine whether entanglement is present or not. In order to transmit information you must also have a classical communication channel.