No.
Consider for example the studies of the Cosmic Background Microwave Radiation. It looked noise-like enough that for many years it was filtered out and ignored as being an artifact. Eventually it was realized there was information in it -- but not until there was a hypothesis about what to look for. Even then, any one picture's CBMR looked like noise and the (apparent) noise of the various images had to be correlated because of the scales involved.
Generally speaking, anything that looks like noise might instead be information.