As part of the current testing, the satellite that is orbiting the earth is using a system of pseudo-random numbers to obscure the data being sent down from the satellite. After requesting the codes and being denied access, the researchers at Cornell decided that a little computing power would do the trick, and set out to break the code.
Slated to begin operation in 2008, the Galileo constellation of satellites is different from the US GPS system in a number of ways, but for consumers in an especially annoying one in that they hope to charge money for consumer access to the feed (and commercial access for that matter). In order to do this, the ESA has decided to encrypt (or scramble) the data and then charge people who sell hardware and software a royalty to decode the data.
As might be expected (and as noted in the article above about the Cornell researchers), this leads to an interesting problem with the legal protections for this data. Since the data is basically just time and/or position data (I'm not certain exactly what's in the transmitted frame), it isn't anything that can be copyrighted under any existing laws. As such, receipt and interpretation of the data is something that is allowed, but inconvenient if it is encoded. Typically, under these circumstances, an organization would use trade secrets to keep the data from unwilling eyes, but the stakes are so high here that people are willing to reverse-engineer the secrets in order to gain access to the data, and so they have. This leaves the ESA in the situation of wondering if they will ever be able to keep the system a source of funds.
Clearly the codes in use today on the test system will differ from those in the final system, but the same (or similar) tactics can be used to decode those and it is unlikely that the secrecy of these codes will remain so for very long.