The issue of copyright has, to a greater or lesser extent, been a problem of Youtube’s from the very beginning. It seems strange, then, that after over 10 years of functioning, it is still nowhere near a state that would be satisfactory both for the largely amateur content producers and for the biggest players on the market. One of the main reasons for this state of affairs is the fact that while the service deals solely with video content, it still means that it encompasses a large number of different industries and forms of intellectual property, each one with their own approach to protecting it. This, in turn, makes the process of coming up with a coherent, site-wide set of policies extremely difficult.
As Youtube serves primarily as a hub for sharing video content, it makes sense that the movie/TV industry has always been the loudest voice against piracy on the website, as it is simply the most obvious victim of illegal activity. This area is also probably the most straightforward in terms of determining what counts as a copyright violation – the distinction between pirated, unaltered footage and transformative content is in most cases rather obvious.
During its first few years, the website was not of much help in trying to fight illegal copies: major full movies and TV shows were easy to find in their entirety – split in 10 minute-long parts and in questionable quality, but still watchable. Without today’s Content ID matching system, the only real way of finding such pirated copies was manual search, therefore simply naming the pirated content more cryptically was a safe bet at ensuring it would stay up for a long time.
A major change came with the introduction of the automated ID system in the early 2010s, forcing the pirates to be somewhat more creative. As in its early form the algorithm was based primarily on image, a new phenomenon of mirrored (flipped along one of vertical edges) copyrighted content emerged. While to a viewer who has not seen the production before, such videos looked absolutely fine, they completely confused the image matching system, making them almost immune to automated takedowns.
In the same vein, videos of brand new releases, recorded by camera in cinemas became another way of tricking the system for two major reasons. Firstly, such releases were likely not yet added to the Youtube database, thus having no original to match the pirated copy to. Secondly, even if they already did, the colors and frame composition recorded by the amateur camera were so off that they were very likely to not be recognized.
Later, however, the automated system largely moved to detecting audio – much less reliant on quality to work properly, assuming no major changes are made. For the most part, this has succeeded in eliminating such problems, solving the piracy problem almost completely – as long as the publishers provide a sample. Apart from some isolated incidents in 2013, major pirated movies on Youtube are currently virtually unheard of. The system, however, is by no means perfect, as the side effects of sometimes being overeager with taking down legitimate content continue to be a massive nuisance to the creators.
Interestingly, the movie industry seems to have realized the medium’s potential in the last few years, as multiple producers/publishers started running channels dedicated to publishing legally and in good quality their older, no longer profitable productions. While finding a movie more recent than ~1970s on those channels is currently almost impossible, it still seems to be a step in the right direction and a sign that the industry is not completely out of touch with today’s culture of sharing.