Amazon’s Audible unit is fighting back against a lawsuit by seven major publishers that claimed Audible’s new automatic caption feature violates their copyrights. In a legal filing last Thursday, Audible argued that the technology is protected by fair use.
The feature uses software to automatically generate text captions as an audio book plays.
Audible was expected to release the feature to users as soon as this month. But Audible tells Ars Technica it will delay a full launch of the technology until litigation wraps up. In the meantime, Audible is offering 150,000 high school students the chance to use the technology with public domain works only.
In their lawsuit, the publishers argued that Audible is effectively distributing an e-book alongside the audio file—something that normally requires a separate license and payment of additional royalties. But Audible disputes that.
“Audible Captions is not a book of any kind, much less a replacement for paper books, e-books, or cross-forma products,” Audible argues in its latest filing.
Audible argues that the technology has substantial public benefits. It can help people with learning disabilities or limited hearing to better understand audiobooks. It can aid people learning a new language. And it can help a growing number of teens who have grown accustomed to watching online videos with captions on—and would benefit from listening to audiobooks the same way.
Audible argues its technology is fair use
Copyright’s fair-use doctrine has enabled a number of key media technologies over the last 40 years. Amazon’s legal filing cites the famous 1984 Supreme Court ruling holding that recording copyrighted TV shows with a video cassette recorder was legal. The company points to a 2015 appeals court ruling upholding Google’s ambitious project to scan millions of copyrighted books for use in its search engine. Audible also mentions a series of cases holding that it was fair use to display thumbnails of copyrighted images in search results.
In each of those cases, the courts sided with the defendants even though they had copied works without the permission of their copyright holders. Fair use considers several factors, including whether the use is “transformative” and whether it will undermine the market for the original work.
Audible argues that it has a strong fair-use argument. The company’s Thursday filing points out that the app never shows more than 20 words on the screen at once, and these words can only be viewed as an audio file is being played. Because the words can only be read at the exact speed the audio plays, it would be a frustrating experience for someone to try to use the captions as a substitute for an e-book.
When I first covered the lawsuit last month, I argued that Audible’s argument “will likely be strengthened by the fact that its app never creates or saves a permanent, full transcript of an audiobook.” It turns out I was wrong.
Audible says its software sends the audio file to Amazon’s servers for transcription, then stores the full transcript on the user’s device. However, Audible takes precautions to make sure the user can only view a few words of the transcript at a time—and only synchronized to the audiobook. These limitations bolster Audible’s fair-use case because the transcript is a poor substitute for an e-book if it can only be viewed at the speed the audio file plays.
The publishers have asked the judge to bar Audible from releasing its technology while the case proceeds in court. The company argues that an injunction isn’t warranted because Audible is likely to win the case—and because Audible can pay cash damages if the publishers ultimately prevail. In practice, the injunction may not matter very much since Audible is already keeping the feature off the market while the lawsuit proceeds.