Apple might have found a new way to clean up the inappropriate talk in your streaming music.
The U.S. Patent and Trademark Office (USPTO) published Apple’s latest patent application on Thursday, describing how the company could manage, replace, and remove “explicit lyrics” in streaming audio. The tech giant’s application describes how the feature would replace the explicit lyrics on the fly with a simple “beep.” Apple posits it could also replace the content with a clean version of the song or silence the vocal track for the moment.
“Generally, in order to avoid unwanted language (e.g., explicit, profane, or otherwise inappropriate language) in songs or other audio the entire song would have to be avoided,” Apple writes in its patent application. The iPhone maker added that its technology would simply remove a portion of the content, so the user would not need to “avoid” the content entirely.
Get Data Sheet, Fortune’s technology newsletter.
The technology, which was previously discovered by Business Insider, relies on metadata, or small pieces of information that would be embedded in the audio file. Apple’s technology would analyze metadata that “describe(s) the location of unwanted audio within an audio stream.” Once it finds such an instance, Apple would then replace the lyrics with something else.
The functionality could have a profound impact on Apple Music, the company’s streaming-audio service. Currently, users need to determine whether streams should include or filter explicit songs. It’s a handy toggle for parents who want to listen to certain songs—but don’t want their kids hearing certain words.
However, in some cases, innuendos can be used, which may not be considered “explicit,” and slip through the cracks. Apple’s patent-pending technology could theoretically identify all of those potential problem areas in a song and replace them.
The trouble is that Apple would likely either need outside help or spend a significant amount of time analyzing songs itself. If the technology relies on metadata, content creators would seemingly need to place some sort of markers into their content to flag Apple’s service to the explicit content. If they didn’t do so, Apple would likely have to scan tracks itself to find potentially offending content. In either case, it appears to be no simple task.
For more about Apple, watch:
To hedge against that, Apple says it could also use “real-time recognition,” which could be used to analyze content on the fly and ensuring it doesn’t have any explicit content. It’s unclear, however, how that content would be analyzed.
The application, filed in September 2014, is one of many inventions that Apple pitches to the USPTO each month. Whether it will eventually make its way to the company’s Apple Music platform, however, is unknown. Like most major companies, Apple often files for patents on a slew of ideas, and in many cases, those technologies never find their way to products.
Read more: Apple Music Just Did in Six Months What Took Spotify Six Years
Still, there is some potential with Apple’s latest invention. Not only would it provide a different way to filter music, but it could also work in other forms. In fact, Apple touts this piece of technology could be used to censor explicit content or curse words in audio books.
