We study the hypothesis of information panspermia assigned recently among possible solutions of the Fermi paradox ("where are the aliens?"). It suggests that the expenses of alien signaling can be significantly reduced, if their messages contain compressed information. To this end we consider universal compression and decoding mechanisms (e.g. the Lempel-Ziv-Welch algorithm) that can reveal non-random structures in compressed bit strings. The efficiency of Kolmogorov stochasticity parameter for detection of non-randomness is illustrated, along with the Zipf's law. The universality of these methods, i.e. independence on data details, can be principal in searching for intelligent messages.