Some law firms will perform copyright searches in an attempt to learn whether an image or work might infringe someone’s copyright. Copyright searches aren’t like other IP searches. Trademark searches are performed to see if a potential mark is likely to be registered or face obstacles during prosecution. Patent searches are performed to see if an invention is novel and nonobvious and can be patented. Copyright searches are a bit different, and often dangerous.
Copyright protection exists from the moment a work is created. Just about everything with a minimal level of creativity has copyright protection. Federal registration of a copyright is another matter – this is a process that must be applied to begin and results in having your copyright federally registered. Copyrights that are registered are relatively easy to search and discover, because the government has a list of them catalogued away. It would be impossible, however, to find all relevant, non-registered, copyrighted images. Most works are never registered for copyright protection, such as school essays, restaurant menu designs, website layouts, childhood drawings, and on and on and on…. All of these things have copyright protection, but they will be almost impossible to find because they are usually never registered. If I draw something and file it away, it has copyright protection, but no one will ever really know about it.
The impossibility of finding them isn’t necessarily a problematic thing. Because copyright infringement generally requires proof of both access and copying, the impossibility of searching and finding a work could correlate with it not being initially accessed, and therefore, not copied. This, too, makes sense: the fact that one kid’s picture of the family, the house, the dog, and the sun looks like another kid’s doesn’t mean they are copies – it simply means they look alike. If the first kid never had access to the other kid’s work, he couldn’t have copied it.
Doing a search opens a can of worms, and that can has the label “access.” If you were to do a search of registered or non-registered copyrighted images and did find something that is similar to your work, you have now accessed that similar work. If you then continue to use your work with your new-found knowledge of the similar work, there is an argument that you have now infringed the similar work’s copyright. So a search doesn’t really help “avoid” infringement: if you copied your image from a work, then you know about the underlying work, and the search is unnecessary. If you didn’t copy your image from a work, then you will only risk acquiring knowledge of a similar work by performing a search and exposing yourself to an infringement claim.